Structurally characterized carboxylic acid modified zirconium alkoxides for the production of zirconium oxide thin films

Article abstract of DOI:10.1016/j.poly.2005.05.005

A series of carboxylic acid (H-ORc) modified zirconium alkoxides (Zr(OR)₄) were synthesized through the reaction of the commercially available [Zr(μ-OPrⁱ)(OPrⁱ)₃(H-OPr ⁱ)]₂ (1, OPrⁱ = OCH(CH₃) ₂) with a series of sterically varied H-ORc, including: formic acid (H-O₂CH or H-OFc), acetic acid (H-O₂CCH₃ or H-OAc), isobutyric acid (H-O₂CCH(CH₃)₂ or H-OPc), trimethyl acetic acid (H-O₂C(CH₃)₃ or H-OBc), and t-butyl acetic acid (H-O₂CCH₂C(CH ₃)₃ or H-ONc) which yielded the following products: Zr₄(μ₄-O)(μ-O)(μ-OFc)₂(μ-OPr ⁱ)₄(OPrⁱ)₆ (2), Zr ₃(μ₃-O)(μ-OAc)₃(OAc)₂(μ- OPrⁱ)₂(OPrⁱ)₃ (3), [Zr ₂(μ-OPc)₂(μ-OPrⁱ)₂(OPr ⁱ)₄]₂ (4), Zr₂(μ-OBc)(μ- OPrⁱ)₂(OPrⁱ)₅(H-OPrⁱ) ? (H-OPrⁱ) (5), Zr₂(μ-ONc)(ONc)(μ-OPr ⁱ)₂(OPrⁱ)₄(H-OPrⁱ) (6).To increase the structural variability of these precursors, we also investigated the H-ORc modifications of the novel Zr₃(μ₃-O) (μ₃-ONep)(μ-ONep)₃(ONep)₆ (7) which was isolated from the reaction between Zr(NMe₂)₄ and 4 H-OCH₂C(CH₃)₃ (H-ONep).The ORc-modified 7 species were isolated as: Zr₃(μ₃-O)(μ-OAc) ₃(μ-ONep)₂(ONep)₅ (8), Zr ₅(μ-O)₃(μ-OPc)₆(μ-ONep) ₂(ONep)₆ (9), [Zr(μ-OPc)(μ-ONep)(ONep) ₂]₂ (10), Zr₅(μ-O)₃(μ-ONc) ₆(μ-ONep)₂(ONep)₆ ? (H-ONep) ? 1/2(C₇H₈) (11). Once fully characterized, these compounds were used to generate thin films of ZrO₂ to investigate the optimal structural aspects that dictate thin film density. It was determined that the majority of these compounds did not yield high quality films; however, the non-condensed species (3-5) did produce clear and continuous ZrO₂ films. From this very limited set of useful precursors, the larger nuclearity species (3) led to films of higher densification.

Full text of DOI:10.1016/j.poly.2005.05.005

Polyhedron 24 (2005) 1727–1738
www.elsevier.com/locate/poly
Structurally characterized carboxylic acid modified zirconium
alkoxides for the production of zirconium oxide thin films
*
Timothy J. Boyle , Leigh Anna M. Ottley, Mark A. Rodriguez
Sandia National Laboratories, Advanced Materials Laboratory, 1001 University Boulevard, S.E., Albuquerque, NM 87106, United States
Received 21 December 2004; accepted 16 May 2005
Available online 14 July 2005
Abstract
A series of carboxylic acid (H-ORc) modified zirconium alkoxides (Zr(OR)4) were synthesized through the reaction of the com-
mercially available [Zr(l-OPri)(OPri)3(H-OPri)]2 (1, OPri = OCH(CH3)2) with a series of sterically varied H-ORc, including: formic
acid (H-O2CH or H-OFc), acetic acid (H-O2CCH3 or H-OAc), isobutyric acid (H-O2CCH(CH3)2 or H-OPc), trimethyl acetic acid
(H-O2C(CH3)3 or H-OBc), and t-butyl acetic acid (H-O2CCH2C(CH3)3 or H-ONc) which yielded the following products: Zr4(l4-
O)(l-O)(l-OFc)2(l-OPri)4(OPri)6 (2), Zr3(l3-O)(l-OAc)3(OAc)2(l-OPri)2(OPri)3 (3), [Zr2(l-OPc)2(l-OPri)2(OPri)4]2 (4), Zr2(l-OBc)-
(l-OPri)2(OPri)5(H-OPri)Æ(H-OPri) (5), Zr2(l-ONc)(ONc)(l-OPri)2(OPri)4(H-OPri) (6).To increase the structural variability of these
precursors, we also investigated the H-ORc modifications of the novel Zr3(l3-O)(l3-ONep)(l-ONep)3(ONep)6 (7) which was iso-
lated from the reaction between Zr(NMe2)4 and 4 H-OCH2C(CH3)3 (H-ONep). The ORc-modified 7 species were isolated as:
Zr3(l3-O)(l-OAc)3(l-ONep)2(ONep)5 (8), Zr5(l-O)3(l-OPc)6(l-ONep)2(ONep)6 (9), [Zr(l-OPc)(l-ONep)(ONep)2]2 (10), Zr5(l-
O)3(l-ONc)6(l-ONep)2(ONep)6 Æ(H-ONep)Æ1/2(C7H8) (11). Once fully characterized, these compounds were used to generate thin
films of ZrO2 to investigate the optimal structural aspects that dictate thin film density. It was determined that the majority of these
compounds did not yield high quality films; however, the non-condensed species (3–5) did produce clear and continuous ZrO2 films.
From this very limited set of useful precursors, the larger nuclearity species (3) led to films of higher densification.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Keywords: Zirconium alkoxides; Zirconium oxide; Neo-pentoxide; Carboxylic acid; Thin films
1. Introduction
through minor changes in the ligand set [11–16]. These
seemingly small changes often alter the overall precursor
structure which ultimately dictate the properties of the
resultant ceramic materials [11–16]. Therefore, in order
to generate the highest quality materials, it is necessary
to understand the structural aspects of these modified
M(OR)x precursors.
As the above ZrO2 based devices become more com-
pact, the ability to produce high quality, dense, ceramic
thin films becomes of increased importance. We are
interested in identifying the structural properties of the
precursors that will generate highly dense ZrO2 thin
films in order to meet this demand. In investigating pre-
viously characterized zirconium alkoxide (Zr(OR)4) pre-
cursors, it was surprising to find that only a few
Zirconium oxide (ZrO2) is an advanced ceramic
material that has found use in many diverse applica-
tions, including catalysis [1,2], DNA extraction [3], lumi-
nescent materials [4], surface supports [5,6], laser
systems [7], gate dielectrics [8,9], optical [10], and elec-
tronic devices. Metal alkoxides (M(OR)x) are excellent
precursors for the production of these advanced ceramic
materials [11–16] mainly due the ease with which their
hydrolysis and condensation rates can be manipulated
*
Corresponding author. Tel.: +1 505 272 7625; fax: +1 505 272
7336.
E-mail address: tjboyle@sandia.gov (T.J. Boyle).
0277-5387/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2005.05.005

1728
T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
homoleptic species had been crystallographically char-
acterized, including [Zr(OR)4 ÆnHOR]x {OR = OCH2-
CH3 (OEt), n = 0, x = 4 [17]; OCH(CH3)2 (OPri), n = 1,
x = 2 [18,19]} while several mixed alkoxy Zr (and Ti)
compounds have been reported [19,20]. Due to this
limited supply of useful precursors, it was necessary to
develop a new family of similarly ligated species that
possessed different structural arrangements. Previously,
we have used a series of structurally characterized car-
boxylic acid (H-ORc, Fig. 1(a)–(e)) modified Ti(OR)4
[OR = OPri; OCH2C(CH3)3 (ONep)], to generate ana-
tase (TiO2) films of varied densities [21,22]. Due to our
success with the congener system, we followed a similar
approach to study ZrO2 thin films.
The widespread utility of ZrO2 thin films [1–12]
makes it of interest to determine the optimal precursors
that will control the densification of the final films. A
search of the literature reveals only a few Zr(OR)4 have
been crystallographically characterized, including
[Zr2(l3-OEt)(l-OEt)(OEt)5]2 [17] (1) [18,19], [Zr(l-OPri)-
(ONep)3(H-ONep)]2 [20], Zr(1,2,5,6-di-O-isopropylid-
ene-a-D-glucfuranose)4(py)2 [23], and [(l-THME)Zr2-
(OPri)5]2 [19] (THME = tris(hydroxymethyl)ethane). The
dinuclear compounds adopt a standard edge-shared
bi-octahedral geometry using two alcohol ligands to
complete their geometry.
Several unusual Zr(OR)4 species, derivatized with a
variety of H-ORc-modifiers have been reported as:
Zr6O2(OPrn)16(O2C-C10H6O)2(PriO-H)2, Zr10O6(OH)4-
(O2C-C6H4OH)8(O2C-C6H4O)8Æ6H-OPr [24], [Zr6(OH)4-
O4(acrylate)10]2(l-acrylate) [25], Zr6(OH)4O4(O2CR)12
(PrOH-Prn) (R = Ph, CMe=CH2) [25], Zr6(l3-O)2(l-
EtOEt)6(l-C3H2O2Me)(C3H2O2Me)4(EtOEt) [26], and
Zr6(l3-O)2(l-OMe)4(l-C3H2O2Me)6(C3H2O2Me)8(EtO-
Et)2 [27]. It is of note that additional ORc-modified
mixed metal alkoxide [26,27] and oxo hydroxy [28–30]
compounds have also been reported.
From these compounds, it is obvious that a system-
atic series of ORc-modified Zr(OR)4 compounds was
not available to initiate a study on the densification of
ZrO2 thin films. Therefore, it was necessary to synthesize
a novel family of ORc-modified Zr(OR)4 precursors that
had similar hydrolysis and condensation routes but were
structurally varied. These compounds would then be
used to generate thin films of ZrO2 to explore the effects
these subtle structural changes had on the density of the
final films. The products we isolated, their characteriza-
tion, and the affect their structures have on the final film
densification are presented.
O
O
O
2. Results and discussion
CH3
CH3
Since [Zr(OPri)4 ÆH-OPri]2 (1, Fig. 1(f)) was commer-
cially available, we initiated our study by modifying the
ligand set of 1 with a series of carboxylic acids (H-ORc),
including formic acid (H-O2CH or H-OFc), acetic acid
(H-O2CCH3 or H-OAc), isobutyric acid (H-O2CCH-
(CH3)2 or H-OPc), trimethyl acetic aid (H-O2CC(CH3)3
or H-OBc), and t-butyl acetic acid (H-O2CCH2C(CH3)3
or H-ONc). Schematic representations of the H-ORc
structures are shown in Fig. 1(a)–(e). For H-ORc-mod-
ified 1 (Fig. 1(f)), the following compounds were charac-
terized as: Zr4(l4-O)(l-O)(l-OFc)2(l-OPri)4(OPri)6 (2),
Zr3(l3-O)(l-OAc)3(OAc)2(l-OPri)2(OPri)3 (3), [Zr2(l-
OPc)2(l-OPri)2(OPri)4]2 (4), Zr2(l-OBc)(l-OPri)2(OPri)5 Æ
(H-OPri)(H-OPri) (5), and Zr2(l-ONc)(ONc)(l-OPri)2-
(OPri)4(H-OPri) (6). Due to its structural variation in
comparison to 1, we furthered our precursor set by mod-
ifying the novel compound [Zr3(l3-O)(l3-ONep)(l-
ONep)3(ONep)6] (7) with the same set of ORc modifiers.
These products were identified as: Zr3(lz-O)(l-OAc)3(l-
ONep)2(ONep)5 (8), Zr5(l-O)3(l-OPc)6(l-ONep)2(O-
Nep)6 (9), [Zr(l-OPc)(l-ONep)(ONep)2]2 (10), and Zr5-
(l-O)3(l-ONc)6(l-ONep)2-(ONep)6(H-ONep)1/2(C7H8)
(11). This new family (2–11) of ORc-modified Zr(OR)4
were characterized by a variety of analytical techniques
and subsequently used to investigate the structural influ-
ences the precursors have on the final densification of
ZrO2 thin films.
C
C
C
HO
C
HO
HO
H
HO
CH3
H
a
b
c
O
CH3
C
O
CH3
CH3
C
C
C
CH3
CH3
HO
C
H2
CH3
d
e
OR
Zr
HOR
R
O
RO
RO
OR
OR
Zr
O
R
HOR
OR
f
Fig. 1. Schematic representation of the various carboxylic acid (H-
ORc) modifiers (a) formic acid (H-O2CH or H-OFc), (b) acetic acid
(H-O2CH3 or H-OAc), (c) iso-butyric acid (H-O2CH(CH3)2 or H-
OPc), (d) trimethyl acetic acid (H-O2C(CH3)3 or H-OBc), (e) t-butyl
acetic acid (H-O2CH2C(CH3)3 or H-ONc), and (f) [Zr(l-OR)(OR)3H-
OR]2 (OR = OCHMe2; 1).

T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
1729
2.1. Syntheses of ORc-modified Zr(OR)4
The synthesis of precursors was initiated with the
commercially available 1 (Fig. 1(f)) and the appropriate
H-ORc (Fig. 1(a)–(e)). After stirring for 12 h, all of the
reactions remained clear and were set aside to allow slow
evaporation of volatile material. If X-ray quality crys-
tals were not isolated in a short time frame, then the
reaction volume was drastically reduced by rotary evap-
oration and the mixture cooled to ꢀ35 ꢁC until crystals
formed. Figs. 2–6 show the plots of the solved structures
of 2–6, respectively. FTIR spectra of 2–6 revealed the
standard OPri peaks with a notable shift of the C@O
stretch of the ORc ligand which is consistent with it
being coordinated to a metal center. For 2 and 4, no
OH stretches were observed; however, for 3, 5, and 6 a
strong broad OH stretch was noted. The resulting pow-
ders were very air-sensitive but elemental analyses were
found to be consistent with the solid state structure.
Previously, we have reported the synthesis of [Zr(l-
OPri)(ONep)3 Æ(H-ONep)]2 [22]. This compound was ini-
tially used in our development of acceptable precursors
for production of ZrO2 thin films; however, the products
isolated were identical to those noted for 7 (vide infra).
Fig. 3. Structure plot of 3. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(1) 2.078(2); Zr(1)–O(2) 2.185(3); Zr(1)–O(4) 2.254(3);
Zr(1)–O(9) 2.212(2); Zr(1)–O(12) 2.295(3); Zr(1)–O(13) 2.265(3);
Zr(1)–O(16) 1.891(2); Zr(2)–O(1) 2.102(2): Zr(2)–O(2) 2.095(3);
Zr(2)–O(3) 2.098(2); Zr(2)–O(5) 2.162(3); Zr(2)–O(6) 2.145(3); Zr(2)–
O(14) 1.926(2). Selected bond angles (ꢁ): O(16)–Zr(1)–O(1) 104.93(11);
O(16)–Zr(1)–O(2) 91.61(11); O(16)–Zr(1)–O(4) 172.25(11); O(16)–
Zr(1)–O(12) 95.56(11); O(16)–Zr(1)–O(9) 91.61(11); O(1)–Zr(1)–O(9)
80.84(9); O(1)–Zr(1)–O(2) 72.24(9); O(1)–Zr(1)–O(4) 81.64(9); O(1)–
Zr(1)–O(13) 145.86(10); O(1)–Zr(1)–O(12) 143.12(9); O(14)–Zr(2)–
O(5) 91.88(11); O(3)–Zr(2)–O(5) 91.65(10); O(2)–Zr(2)–O(5)
87.22(10); O(1)–Zr(2)–O(5) 87.15(10); O(6)–Zr(2)–O(5) 175.53(10);
O(14)–Zr(2)–O(3) 105.92(11); O(14)–Zr(2)–O(2) 105.70(11); O(14)–
Zr(2)–O(1) 179.08(11); O(14)–Zr(2)–O(6) 92.59(11).
This is not unexpected since the less sterically hindering
OPri ligand would be the most susceptible to exchange
and hydrolysis, leaving only ONep ligands for structural
influences. Therefore, the full series for this precursor
was not pursued.
From an in situ generated amide, the ONep deriva-
tive of Zr was isolated as 7 (Eq. (1)), an off white pow-
der. This compound was selected for further
modification since the trinuclear central core should al-
low for a wide range of structural variations in compar-
ison to the OPri derivatives. The OFc reaction yielded a
precipitate that could not be re-dissolved. All of the
remaining H-ORc modifications were undertaken using
toluene as the solvent with no color changes or precipi-
tates observed. Figs. 7–11 show the plots of structure
plots of 7 and the ORc modified species 8–11. The FT-
IR spectra of these compounds revealed standard ONep
stretches and bends intermixed with stretches that indi-
cated the ORc ligand was coordinated to the Zr metal
center. No –OH stretches were observed. The elemental
analysis of 8–11 proved to be inconsistent with the solid
Fig. 2. Structure plot of 2. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(1) 2.178(2); Zr(1)–O(2) 2.234(2); Zr(1)–O(4) 2.154(2);
Zr(1)–O(3) 1.932(3); Zr(1)–O(5) 1.914(3); Zr(1)–O(6) 2.180(3). Selected
bond angles (ꢁ): O(5)–Zr(1)–O(6) 90.82(10); O(5)–Zr(1)–O(4)
100.93(10); O(5)–Zr(1)–O(2) 168.75(10); O(5)–Zr(1)–O(1) 96.22(10);
O(5)–Zr(1)–O(3) 100.25(11); O(6)–Zr(1)–O(3) 75.38(7); O(4)–Zr(1)–
O(3) 83.70(6); O(2)–Zr(1)–O(3) 88.37(10); O(1)–Zr(1)–O(3) 163.53(10).

1730
T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
Fig. 4. Structure plot of 4. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
Fig. 6. Structure plot of 6. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(1) 2.173(4); Zr(1)–O(2) 1.933(4); Zr(1)–O(3) 1.932(4);
Zr(1)–O(4) 2.109(4); Zr(1)–O(5) 2.181(4); Zr(1)–O(6) 2.181(4). Selected
bond angles (ꢁ): O(3)–Zr(1)–O(2) 99.31(18); O(3)–Zr(1)–O(4)
96.27(17); O(3)–Zr(1)–O(1) 88.54(16); O(3)–Zr(1)–O(6) 94.15(16);
O(2)–Zr(1)–O(5) 95.04(17); O(4)–Zr(1)–O(5) 90.78(15); O(1)–Zr(1)–
O(5) 83.14(15); O(6)–Zr(1)–O(5) 70.89(15).
˚
(A): Zr(1)–O(1) 2.151(6); Zr(1)–O(2) 2.162(6); Zr(1)–O(3) 2.155(6);
Zr(1)–O(5) 2.195(6); Zr(1)–O(6) 1.924(6); Zr(1)–O(7) 1.921(6); Zr(2)–
O(1) 2.176(6); Zr(2)–O(2) 2.192(6); Zr(2)–O(4) 2.183(6); Zr(2)–O(8)
1.925(6); Zr(2)–O(9) 1.926(6); Zr(2)–O(10) 2.074(6). O(11)–H(50)
2.019; O(5)–H(50) 0.85. Selected bond angles (ꢁ): O(7)–Zr(1)–O(6)
99.6(3); O(7)–Zr(1)–O(2) 165.2(2); O(7)–Zr(1)–O(1) 92.2(2); O(7)–
Zr(1)–O(3) 93.3(3); O(7)–Zr(1)–O(5) 94.3(3); O(8)–Zr(2)–O(2)
165.2(2); O(9)–Zr(2)–O(2) 92.7(3); O(10)–Zr(2)–O(2) 90.4(3); O(1)–
Zr(2)–O(2) 71.5(2); O(4)–Zr(2)–O(2) 81.9(2); C(22)–O(11)–H(50)
88.25(6); Zr(1)–O(5)–H(50) 109.25(6).
state structures; however, when a molecule of solvent
(i.e., toluene) was included the analyses fell within the
acceptable range of error.
tol
ZrCl4 þ 4LiNMe2 ! ZrðNMe2Þ4
ꢀ4LiCl
þ4HONep
1=3Zr ðOÞðONepÞ
ð1Þ
ꢀꢀꢀꢀ!
3
10
2.2. Single crystal X-ray structures of ORc-modified
Zr(OR)4
In order to determine the solid state structures of
2–11, single crystal X-ray analysis were undertaken.
The crystal of 2–11 were found to be very air sensitive
and therefore were handled under inert conditions to
minimize exposure to atmospheric conditions while the
crystals were being mounted. Discussions concerning
the structural aspects of 1–11 are detailed below.
Fig. 5. Structure plot of 5. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(1) 2.058(4); Zr(1)–O(2) 1.946(4); Zr(1)–O(3) 1.941(4);
Zr(1)–O(4) 2.263(4); Zr(1)–O(5)2.202(4); Zr(1)–O(6) 2.192(4); O(11)–
H(1) 0.8400; O(7)–H(1) 0.845(18); O(1)–H(2) 1.862; O(11)–H(1) 1.740.
Selected bond angles (ꢁ): O(3)–Zr(1)–O(2) 99.94(11); O(3)–Zr(1)–O(1)
97.09(10); (3)–Zr(1)–O(6) 161.42(10); O(3)–Zr(1)–O(4) 87.21(9); O(3)–
Zr(1)–O(4) 87.21(9); O(2)–Zr(1)–O(4) 88.24(9); O(1)–Zr(1)–O(4)
172.62(8); O(6)–Zr(1)–O(4) 82.48(8); O(5)–Zr(1)–O(4) 82.57(8); O(1)–
H(2)–O(11) 167.34(2); O(7)–H(1)–O(11) 175.56(3).
2.2.1. Single crystal X-ray structures of OPri derivatives
The commercially available 1 is shown in Fig. 1(f)
and several reports on its structure are available
[18,19]. Each Zr metal center is octahedrally (Oh) bound
by two l-OPri, three terminal OPri, and one coordinated

T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
1731
Fig. 8. Structure plot of 8. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
Fig. 7. Structure plot of 7. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(5) 1.919(4); Zr(1)–O(1) 2.063(4); Zr(1)–O(2) 2.125(4);
˚
(A): Zr(1)–O(8) 1.903(5); Zr(1)–O(6) 1.925(4); Zr(1)–O(1) 2.082(3);
Zr(1)–O(8) 2.130(4); Zr(1)–O(4) 2.141(4); Zr(1)–O(6) 2.141(4). Zr(3)–
O(13) 1.916(4); Zr(3)–O(14) 1.948(4); Zr(3)–O(3) 2.058(4); Zr(3)–O(12)
2.176(4); Zr(3)–O(2) 2.204(4); Zr(3)–O(3) 2.291(4). Selected bond
angles (ꢁ): O(5)–Zr(1)–O(1) 179.45(18); O(5)–Zr(1)–O(2) 104.30(18);
O(1)–Zr(1)–O(2) 75.15(16); O(5)–Zr(1)–O(8) 104.75(17); O(1)–Zr(1)–
O(8) 75.80(16); O(2)–Zr(1)–O(8) 150.86(16); O(5)–Zr(1)–O(4)
93.02(18); O(1)–Zr(1)–O(4) 86.99(17); O(2)–Zr(1)–O(4) 86.41(17);
O(8)–Zr(1)–O(4) 89.78(17); O(5)–Zr(1)–O(6) 94.52(18); O(1)–Zr(1)–
O(6) 85.50(17); O(2)–Zr(1)–O(6) 94.28(17); O(8)–Zr(1)–O(6) 85.74(16);
O(4)–Zr(1)–O(6) 172.01(17).
Zr(1)–O(5) 2.164(3); Zr(1)–O(2) 2.167(4); Zr(1)–O(7) 2.196(7); Zr(1)–
O(7) 2.742(6);. Selected bond angles (ꢁ): O(8)–Zr(1)–O(6) 101.9(3);
O(8)–Zr(1)–O(1) 99.0(2); O(8)–Zr(1)–O(5) 105.90(19); O(8)–Zr(1)–
O(2) 100.7(2); O(8)–Zr(1)–O(7) 165.2(2); O(6)–Zr(1)–O(7) 89.4(2);
O(1)–Zr(1)–O(7) 70.4(2); O(5)–Zr(1)–O(7) 81.6(2); O(2)–Zr(1)–O(7)
67.1(2).
H-OPri molecule. This molecule is very robust and has
been used as a sterically hindering ligand for other metal
centers [31–35]. Therefore, it was expected that we
would observe structures that possessed multiplets of
the dinuclear subunit, substituted with ORc ligands.
The OFc derivative of 1 was isolated as the tetranu-
clear species 2 (Fig. 2), wherein, the four Zr atoms are
tetrahedrally (Td) arranged around a l4-O. Another
oxide is present as l-O atom linking two of the four
Zr atoms with the remaining ligands asymmetrically ar-
ranged around this central core: four l-OPri, six termi-
nal OPri, and two l-OFc ligands. Both OFc ligands
bridge two different Zr atoms. This arrangement was
unexpected because compound 2 is structurally identical
to the OFc derivative of Ti(OPri)4 (‘‘Ti-OPri/OFc’’) de-
scribed as a face-shared corner-missing cube structure,
[Ti4(l4-O)(l-O)(l-OFc)(l-OPri)4(OPri)6] [22]. The basic
core structure of 2 appears to have significantly altered
the dinuclear starting material 1 but 2 can be thought
of as two of the dinuclear species of 1 bound perpendic-
ularly by OFc and O ligands.
a l3-O with a symmetric distribution of a single
l-OAc, two l-OPri, and a terminal OPri for each metal
center. The unique Zr metal center bridges to each of the
PBP-coordinated Zr metal centers using a l-OAc ligand.
Finishing the PBP coordination of these Zr metal cen-
ters is a chelating OAc ligand. Based on the previous re-
ported stability of 1, it was surprising that the central
core was so easily disrupted by HOAc forming a struc-
ture similar to ‘‘Ti-ONep/OFc’’ complex which gener-
ated the highest dense films of TiO2 [22].
For compound 4, even though the steric bulk of the
pendant chain of OPc (CH(CH3)2) is greater than the
OAc (CH3), an increase in nuclearity was observed
(see Fig. 4) possibly due to the lack of oxo formation.
The structure possesses a C2h symmetry where the two
sub-moieties are linked by l-OPc ligands. These smaller
units consist of two Zr metal centers bridged by two
OPri, one l-OPc ligand, and two terminal OPri ligands.
This molecule can be thought of as linking two OPc
substituted moieties of 1 (multiplet of the subunit).
There are no similar Ti based structures known.
By increasing the steric bulk of the ORc ligand from a
–H (OFc) to a –CH3 (OAc) group, the nuclearity of the
resulting product 3 was reduced to a trinuclear species
(Fig. 3). The molecule possesses a C2 center of symmetry
with two Zr atoms adopting a pentagonal bipyramidal
(PBP) geometry whereas the unique Zr atom adopted
an Oh geometry. The three Zr metal centers surround
Introduction of the most sterically demanding ligand
investigated, OBc, led to the isolation of the simple
substituted complex 5, shown in Fig. 5. This molecule
consists of two Oh-coordinated Zr atoms bridged by

1732
T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
Fig. 9. Structure plot of 9. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
Fig. 11. Structure plot of 11. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(17) 2.020(4); Zr(1)–O(19) 2.033(4); Zr(1)–O(7) 2.180(4);
˚
(A): Zr(1)–O(7) 1.914(5); Zr(1)–O(1) 1.950(5); Zr(1)–O(5) 2.101(4);
Zr(1)–O(2) 2.182(4); Zr(1)–O(16) 2.183(4); Zr(1)–O(10) 2.191(4);
Zr(1)–O(5) 2.273(4). Selected bond angles (ꢁ): O(17)–Zr(1)–O(19)
147.13(16); O(17)–Zr(1)–O(7) 92.17(15); O(19)–Zr(1)–O(7) 82.84(15);
O(17)–Zr(1)–O(2) 82.56(16); O(19)–Zr(1)–O(2) 126.32(15); O(7)–
Zr(1)–O(2) 74.25(17); O(17)–Zr(1)–O(16) 125.397(16); O(19)–Zr(1)–
O(16) 82.26(16); O(7)–Zr(1)–O(16) 123.44(16); O(2)–Zr(1)–O(16)
72.01(17); O(17)–Zr(1)–O(10) 82.18(15); O(19)–Zr(1)–O(10) 123.31(17);
O(16)–Zr(1)–O(10) 74.17(16); O(17)–Zr(1)–O(5) 73.73(14); O(19)–
Zr(1)–O(5) 73.40(14); O(7)–Zr(1)–O(5) 79.94(14); O(2)–Zr(1)–O(5)
144.03(15); O(16)–Zr(1)–O(5) 143.96(15); O(10)–Zr(1)–O(5) 80.27(14).
Zr(1)–O(4) 2.136(5); Zr(1)–O(6) 2.174(5); Zr(1)–O(11) 2.272(5).
Selected bond angles (ꢁ): O(7)–Zr(1)–O(1) 98.6(2); O(7)–Zr(1)–O(5)
97.33(18); O(7)–Zr(1)–O(4) 93.7(2); O(7)–Zr(1)–O(6) 101.24(19); O(7)–
Zr(1)–O(11) 177.2(2); O(1)–Zr(1)–O(11) 81.46(19); O(5)–Zr(1)–O(11)
82.87(17); O(4)–Zr(1)–O(11) 83.58(19); O(2)–Zr(1)–O(11) 81.48(18).
two l-OPri and one l-OBc ligands. The coordination is
completed by three OPri ligands. Due to charge balance,
one of these ligands must be an H-OPri with an addi-
tional H-OPri was also found in the unit cell. From
bond distances and angles, it appears that there is
H-bonding between the unit cell H-OPri, free H-OPri,
and terminal OPri. Compound 6 was dinuclear as well
with a similar arrangement with one l-ONc, one termi-
nal ONc, two l-OPri, four OPri, and a bound H-OPri. A
structure plot of 6 is shown in Fig. 6. The H-OPri also
appears to be H-bound to the terminal acid. Both 5
and 6 are similar to what was observed for the ‘‘Ti-
OPri/OBc’’ species [22] and again show retention of
the Zr dinuclear subunit.
2.2.2. Single crystal X-ray structures of ONep derivatives
The structure of 7 is a symmetric trinuclear species with
l3-O and l3-ONep ligands (see Fig. 7). The three Zr atoms
are bridged by a l-ONep ligand and the final Oh coordi-
nation is filled by terminal ONep ligands. In contrast,
for the Ti-ONep system, generated through the same reac-
tion pathway, a halide was retained [36]. The variation in
the structure of 7 in comparison to 1 was of interest to fur-
ther modify with ORc ligands to increase the number of
precursors for thin film investigation. Modification with
H-OFc, yielded precipitates that could not be re-dissolved
and X-ray quality crystals from the H-OBc derivatives
could not be isolated in our hands.
Fig. 10. Structure plot of 10. Thermal ellipsoids of heavy atoms are
drawn at the 30% level, hydrogen atoms are omitted, and carbon
atoms are shown as stick and ball for clarity. Selected bond distances
˚
(A): Zr(1)–O(4) 1.909(6); Zr(1)–O(3) 1.939(6); Zr(1)–O(2) 2.163(6);
Zr(1)–O(5) 2.182(6); Zr(1)–O(2) 2.184(6); Zr(1)–O(1) 2.201(6). Selected
bond angles (ꢁ): O(4)–Zr(1)–O(3) 96.3(2); O(4)–Zr(1)–O(2) 168.1(2);
O(3)–Zr(1)–O(2) 92.9(2); O(4)–Zr(1)–O(2) 94.4(2); O(3)–Zr(1)–O(2)
165.9(2); O(2)–Zr(1)–O(2) 78.7(2); O(5)–Zr(1)–O(2) 81.4(2); O(4)–
Zr(1)–O(1) 107.2(2); O(3)–Zr(1)–O(1) 92.0(2); O(2)–Zr(1)–O(1)
80.8(2); O(5)–Zr(1)–O(14) 150.8(2).
Modification of 7 with H-OAc led to the trinuclear
species 8 (Fig. 8). The central core of 8 is similar to 7

T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
1733
with the l3-O bridging the three Oh-coordinated Zr me-
tal centers. Two l-ONep and two l-OAc ligands bridge
to Zr(1) from Zr(2) and Zr(3). The latter metal centers
are bridged by a single l-OAc ligand. Zr(1) finishes its
coordination by binding one terminal ONep ligand
whereas Zr(2) and Zr(3) possess two terminal ONep li-
gands. From this structure, the general constructs of 7
were found to be stable with simple OR–ORc exchanges.
The constructs of compound 8 are similar to 3 and the
‘‘Ti-OPri/OFc’’ complex [22].
For the OPc derivative of 7, two species were isolated,
9 and 10. Compound 9 appears to be the major product
based on the elemental analysis and numerous single
crystal experiments. Compound 9 is a ladder-like struc-
ture containing five Zr atoms (see Fig. 9). The molecule
contains a C2 center of symmetry that creates three types
of Zr atoms (ratios of 2:1:2). The unique Zr has no ter-
minal ONep ligands but uses four l-OPc and three l3-O
ligands to form a PBP geometry around the 7-coordi-
nated Zr atoms. Two Oh-coordinated Zr atoms have
one terminal ONep ligand and to l-OPc, two l3-O,
and one l-ONep set of ligands. The final two Zr are
Oh bound with two OPc, two terminal ONep, one
l-ONep, and one l3-O ligand sets. Compound 9 is
unique in the previously identified Group 4 OR-ORc
structure types [22] and reveals the disruption of the tri-
nuclear 7. Another structure was isolated from this reac-
tion mixture, 10 (Fig. 10). This compound is a simple
dinuclear species with two l-ONep, two l-OPc, and four
terminal ONep ligands filling the Oh geometry of the
two Zr atoms; however, 10 is similar to the structure
noted for the ‘‘Ti-ONep/OBc’’ species [22]. Compound
11 is structurally similar to 9 with ONc ligands replacing
the OPc ligands and is shown in Fig. 11.
and 6, a 1:1 ratio of ORc carbon (R-CO2) resonances
should be observed, for 3, a 2:2:1 ratio and for 5 only
one resonance should be observed. For 2, only one set
of broad 13C resonance was noted for the disparate OPri
ligands and a very weak but broad OFc resonance. The
spectrum of 3 also showed numerous broad peaks for
the OPri peaks but only a single broad OAc resonance.
Compounds 4 and 5 did not yield high quality spectra
due to the low solubility of these compounds. The spec-
trum of 6 revealed the presence of two types of ONc li-
gands and multiple resonances for the OPri ligands.
The ONep derivatives (8, 9, and 11) are expected to
yield a 2:1 ratio of ligand resonances (9 and 11 may also
display a 2:2:1 ratio). Again, due to the complexity of
1
the H NMR spectrum for each sample, information
concerning the nuclearity of these compounds could
not be discerned; however, the appropriate ligand reso-
nances for the various samples were noted. From the 13
C
NMR spectra, only a single resonance for 8 was ob-
served. With rapid ONep ligand exchange the unequiv-
alent ORc ligands could easily become equivalent and
therefore a singlet is not unexpected. For 9, the expected
three resonances were noted. In contrast, compound 11
revealed only a single broad resonance for the ONc
derivative; however, the alkyl peaks were also broad.
These results indicated that a great deal of ligand scram-
bling is occurring.
Although solution behavior was difficult to establish
because of low compound solubility and observed ligand
scrambling, we assumed, based on the similarity of
structures between the Ti and Zr systems, that similar
solution behavior should occur. For the Ti system, li-
gand scrambling did not alter the general constructs
[22] and this concept was applied to the Zr system de-
scribed here.
2.3. Solution behavior of ORc-modified Zr(OR)4 species
2.4. Thin film behavior of ORc-modified Zr(OR)4
Since the precursor will necessarily be dissolved in a
solvent prior to spin-cast deposition, it is important, if
possible, to understand their solution behavior. To
investigate whether the solid state structure were re-
tained in solution, NMR studies were undertaken by
It was of interest to determine how the various struc-
tural changes of these compounds affected the densifica-
tion of the final ZrO2 thin films. For the Ti system, the
oxo species appeared to assist in holding the compounds
together during hydrolysis and trinuclear species with
some oxo ligands present were found to be the best pre-
cursors to generate highly dense anatase thin films [22].
Due to the low solubility of the Zr compounds it was
difficult to generate solutions that produced films with-
out cracks or large defects. The defects formed during
film processing are most likely due to the increased steric
hindrance of the ONep ligand preventing controlled
cross-linking resulting from exposure to ambient humid-
ity to form the necessary extended Zr–O–Zr network of
a film. Instead, upon condensation the limited interac-
tion would cause closo-structures to form, negating ex-
tended networks causing cracks. This general
argument also can be used to explain the poor properties
1
dissolving crystals of 2–11 in toluene-d8. The H NMR
spectra did not reveal a great deal of information
concerning the structural aspects of 2–6 due to the
numerous, overlapping peaks that were present. As pre-
viously reported for the Ti system [22], rapid ligand-
exchange processes are common for these types of
species and would explain the complicated NMR spec-
tra. For each sample resonances were clearly present
for OPri and the appropriate ORc ligands. In an attempt
to further clarify the solution behavior, 13C NMR spec-
tra were obtained for 2–6. The resonances associated
with the ORc ligands should yield the most information
concerning the solution behavior of these compounds. If
the solid state structures are retained in solution, for 2,4,

1734
T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
of all of the ONep derivatives in comparison to the
smaller OPri species.
4.1. Physical measurements
Acceptable films were generated from solutions of
3–5. In contrast to the characteristics of the optimal pre-
cursors for generation of dense antase films, none of
the Zr compounds that made the best films possess
oxo moieties but are simply the substituted compounds.
From the films that could be analyzed, it was discovered
that by increasing the steric bulk of the pendant ORc li-
gand increased the film thickness which ranged from
FT-IR data were obtained on a Bruker Vector 22
MIR Spectrometer using KBr pellets under an atmo-
sphere of flowing nitrogen. Elemental analyses were per-
formed on a Perkin–Elmer 2400 CHN-S/O Elemental
Analyzer. All NMR samples were handled and stored
under an argon atmosphere and re-dissolved in tolu-
ene-d8 at saturated solution concentrations immediately
prior to use. All solution spectra were obtained on a
Bruker DRX400 spectrometer at 399.8 and 100.1 MHz
for 1H and 13C experiments, respectively. A 5 mm
˚
2500 to 2700 A. However, due to the poor quality, addi-
tional ligand manipulation will be necessary to fully ex-
plore higher density films. In general, it appears that the
same trends observed for the anatase system did not ap-
ply to the zirconium oxide thin films. This is most likely
due to the different hydrolysis rates of the two metal cat-
ions. Efforts to improve the solubility of the Zr precur-
sors are underway in order to garner a larger set of
useful precursors necessary to adequately explore the
densification of zirconium oxide thin films.
1
broadband probe was used for all experiments. The H
NMR spectra were obtained using a direct single pulse
excitation, with a 10 s recycle delay and 8 signal aver-
ages. The 13C{1H} NMR spectra were obtained using
1
a WALTZ-16 composite pulse H decoupling, a 5 s re-
cycle delay and a p/4 pulse excitation.
4.2. General synthesis of ORc-modified Zr(OR)4
The appropriate Zr(OR)4 (1 or 7) was dissolved in
toluene and then added via pipette to a stirring mixture
of the desired carboxylic acid (H-ORc, Fig. 1) already
dissolved in toluene. After 12 h, the reaction was al-
lowed sit to slowly evaporate the volatile material until
crystals formed. If crystals did not form, the reaction
mixture was further concentrated by rotary evaporation
and then placed into a freezer (ꢀ35 ꢁC) until crystals
formed. Scale and analytical data are presented for each
compound (vide infra).
3. Summary and conclusions
We have synthesized and characterized a series of
ORc-modified Zr(OR)4 species using commercially
available 1 and specialty precursor 7. The molecules of
the OPri derivatives range from simple substitutions of
ORc for OR groups resulting in tetranuclear (4) and
dinuclear (5 and 6) species to partially hydrolyzed trinu-
clear species (2 and 3). For the derivatives of 7, the
majority of species had an oxo group present but this
may be a reflection of the starting materials composi-
tion. The nuclearity of these species was either tri- (8)
or pentanuclear (9 and 11). Surprisingly, the previously
unobserved structure type was found for two acid
modifiers in this system. Of these species, three (3–5) gen-
erated acceptable quality films which could be monitored
for film thickness and densification. The remainder
produced cracked films. For 3–5, the larger pendant
ORc chains generated the thickest films.
4.2.1. Zr4(O)2(OFc)2(OPri)10 (2)
Used 1 (0.500 g, 1.29 mmol), H-OFc (0.0590 g, 1.29
mmol) and ꢁ10 mL of toluene. Yield 0.390 g (56.1%).
FT-IR (KBr, cmꢀ1) 2969(s), 2931(s), 2864(s,sh), 2626(m),
2364(w), 2345(w), 1590(s), 1463(s), 1375(s), 1338(s,sh),
1167(s), 1051(w,sh), 1018(s), 955(s), 847(s). Elemental
Anal. Calc. for Zr4O16C32H72: C, 35.66; H, 6.73. Found:
C, 35.63; H, 6.36%.
4.2.2. Zr3(O)(OAc)5(OPri)5 (3)
Used 1 (0.500 g, 1.29 mmol), H-OAc (0.0770 g,
1.29 mmol) and ꢁ10 mL of toluene, heated for 10 min
until solution went clear. Yield 0.350 g (46.3%). FT-IR
4. Experimental
All compounds described below were handled with
rigorous exclusion of air and water using standard
Schlenk line and glove box techniques. All solvents were
used as received from Aldrich in sure seal bottles.
H-OFc (Aldrich) and H-OAc (Fisher) were dried and
freshly distilled, immediately prior to use. The following
compounds were stored under argon upon receipt (Al-
drich) and used without further purification: H-ONep,
H-OPc, H-OBc, H-ONc, 1, ZrCl4, and LiNMe2. Sam-
ples for all analyses were prepared from dried crystalline
material.
(KBr, cmꢀ1
) 2969(s), 2932(m), 2863(w), 2625(w),
1548(s), 1450(s), 1375(m), 1363(m,sh), 1337(w),
1168(s), 1049(sh), 1017(s), 947(s), 847(m), 826(m,sh),
653(m), 617(m), 554(s), 461(s). Elemental Anal. Calc.
for Zr3O16C25H50: C, 34.11; H, 5.72. Found: C, 34.04;
H, 5.62%.
4.2.3. [Zr2(OPc)2(OPri)6]2 (4)
Used
1 (0.500 g, 1.29 mmol), H-OPc (0.113 g,
1.29 mmol) and ꢁ10 mL of toluene, heated for 10 min
until solution went clear. Yield 0.370 g (40.4%). FT-IR

T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
1735
(KBr, cmꢀ1) 2970(s), 2932(m), 2868(m), 2624(w), 2367(w),
2345(w), 1578(s), 1561(s,sh), 1510(w), 1476(m), 1439(s),
1375(s), 1362(s,sh), 1338(m), 1301(m), 1262(w), 1169(s),
1099(sh), 1021(s), 1009(sh), 974(m), 847(m), 824(m),
775(w), 654(m), 542(s), 462(s). Elemental Anal. Calc. for
Zr2O10C26H56: C, 43.91; H, 7.94. Found: C, 44.16; H,
7.61%.
4.2.7. Zr3(O)(OAc)3(ONep)7 (8)
Used 7 (0.500 g, 0.431 mmol), H-OAc (0.0780 g,
0.431 mmol), and ꢁ10 mL of toluene. Yield 0.300 g
(64.7%). FT-IR (KBr, cmꢀ1) 2953(s), 2906(w), 2867(s),
2834(s,sh), 2704(w), 2362(w), 2344(w), 1572(s), 1479(s),
1458(s), 1396(s), 1362(s), 1296(w). 1260(w), 1143(s),
1120(s,sh), 1060(m), 1023(s), 994(m), 936(m), 921(m),
751(m), 632(s), 588(m), 538(s), 453(s). Elemental Anal.
Calc. for Zr3O14C41H86: C, 45.73; H, 8.04. Found: C,
49.73; H, 8.50%.
4.2.4. Zr2(OBc)(OPri)7(H-OPri)(H-OPri) (5)
Used
1 (0.500 g, 1.29 mmol), H-OBc (0.132 g,
1.29 mmol) and ꢁ10 mL of toluene. Yield 0.400 g
(38.1%). FT-IR (KBr, cmꢀ1) 2969(s), 2930(m), 2865(w),
2625(w), 2346(w), 1583(s,sh), 1561(s), 1509(w), 1485(s),
1460(m,sh), 1431(s), 1407(w,sh), 1374(s), 1361(s), 1339(m),
1229(m), 1171(s), 1023(s), 945(s), 898(w), 847(m), 822(m),
791(w), 601(s), 561(s), 545(s), 461(s), 432(w,sh), 407(m).
Elemental Anal. Calc. for C32H74O11Zr2: C, 47.02; H,
9.12. Found: C, 46.82; H, 8.87%.
4.2.8. Zr5(O)3(OPc)6(ONep)8 (9)
Used 7 (0.500 g, 0.431 mmol), H-OPc (0.114 g,
0.431 mmol) and ꢁ10 ml of toluene, heated for 10 min
until solution went clear. Yield 0.198 g (86.8%). FT-IR
(KBr,cmꢀ1) 2953(s), 2868(s), 2796(w), 2701(m), 2362(m),
2344(m), 1562(s), 1523(w,sh), 1476(s), 1431(s), 1395
(w,sh), 1678(w), 1368(m), 1298(s), 1266(w,sh), 1128(s),
1088(w,sh), 1055(w,sh), 1037(m), 1026(m), 1010(s), 964
(w), 935(s), 901(m), 861(m), 840(w), 763(w,sh), 750(m),
623(s), 579(s), 527(s), 456(s). Elemental Anal. Calc. for
Zr5O23C64H130: C, 44.59; H, 7.60. Found: C, 46.31; H,
7.12%.
4.2.5. Zr2(l-ONc)(H-ONc)(OPri)7 (6)
Used 1 (0.500 g, 1.29 mmol), H-ONc (0.150 g, 1.29
mmol) and ꢁ10 mL of toluene. Yield 0.430 g (40.2%).
FT-IR (KBr, cmꢀ1) 2966(s), 2867(s,sh), 2625(m), 1561(s),
1459(s), 1337(s,sh), 1307(m), 1277(m), 1237(m), 1168(s),
1022(s), 948(s), 847(m), 821(m), 726(m), 628(sh),
548(s), 462(s), 411(s, sh). Elemental Anal. Calc. for
Zr2O11C33H72: C, 47.91; H, 8.77. Found: C, 47.56; H,
8.43%.
4.2.9. Zr5(O)3(ONc)6(ONep)8(H-ONep)1/2(C7H8)
(11)
Used 7 (0.500 g, 0.431 mmol), H-ONc (0.150 g,
0.431 mmol) and ꢁ10 mL of toluene. Yield 0.390 g
(74.4%). FT-IR (KBr, cmꢀ1) 2955(s), 2907(m), 2870(w),
2719(w), 2370(w), 2345(w), 1708(s), 1612(m,sh),
1571(s), 1528(w,sh), 1459(s), 1414(s), 1367(s), 1307(s),
4.2.6. Zr3(O)(ONep)10 (7)
Zr(NMe2)4 was synthesized from the reaction of
ZrCl4 (1.00 g, 4.29 mmol) and LiNMe2 (0.876 g,
17.1 mmol) in THF (ꢁ8 mL). The precipitate was re-
moved by centrifugation, the soluble fraction dried, ex-
tracted with hexanes, and dried by rotary evaporation.
The Zr(NMe2)4 (0.869 g, 3.24 mmol) powder was re-dis-
solved in toluene (ꢁ5 mL), H-ONep (1.28 g, 14.6 mmol)
in toluene (ꢁ5 mL) was added and the reaction was al-
lowed to stir. After 8 h, the resultant mixture was
warmed (ꢁ60 ꢁC) for 1 h, allowed to cool to room tem-
perature and all of the volatile material removed in va-
cuo. The resultant white powder was dissolved in a
warmed solution of hexanes and toluene (3:1) and
allowed to cool to room temperature. X-ray quality
crystals were isolated. Yield 1.06 g (63.8%). FT-IR
(KBr, cmꢀ1) 2952(s), 2866(s,sh), 2701(w,sh), 2363(m),
2345(m), 1479(s), 1467(m,sh), 1395(s), 1362(s),
1259(m), 1144(s), 1056(m), 1026(w,sh), 958(w), 901(w),
1276(s),
1235(s),
1198(m),
1145(m),
1047(m),
1035(w,sh), 978(m), 936(w), 910(m), 811(m), 757(m),
726(m), 691(m), 640(s), 593(s), 500(m,sh), 449(m),
411(s). Elemental Anal. Calc. for C84.5H170O24Zr5: C,
50.09; H, 8.46 and for solvent free: Calc. for
C75H155O23Zr5: C, 47.89; H, 8.30. Found: C, 48.12; H,
8.77%.
4.3. General X-ray crystal structure information [37]
Each crystal was mounted onto a thin glass fiber from
a pool of Fluorolubee and immediately placed under a
liquid N2 boil-off stream, on a Bruker AXS diffractom-
eter. The radiation used was graphite monochromatized
˚
Mo Ka radiation (k = 0.7107 A). The lattice parameters
were optimized from a least-squares calculation on care-
fully centered reflections. Lattice determination and
data collection were carried out using SMART version
5.054 software. Data reduction was performed using
SAINT version 6.01 software. The structure refinement
was performed using XSHELL 3.0 software. The data were
corrected for absorption using the SADABS program with-
in the SAINT software package.
1
617(m), 517(m), 455(m). H NMR (400 MHz, toluene-
d8) d 3.84 (2H, s, OCH2C(CH3)3), 1.02, 0.97 (12.8H, s,
OCH2C(CH3)3). 13C{1H} NMR (100.1 MHz, toluene-
d6) d 77.7 (OCH2C(CH3)3), 24.7, (OCH2C(CH3)3),
28.1, 27.3 (OCH2C(CH3)3). Elemental Anal. Calc. for
C50H110O11Zr3: C, 51.72; H, 9.54. Found: C, 52.20; H,
9.54%.
Each structure was solved using direct methods. This
procedure yielded the heavy atoms, along with a number

1736
T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
of the C atoms. Subsequent Fourier synthesis yielded the
remaining C atom positions. The crystal structures of me-
tal alkoxides are often plagued by disorder in the pendant
chain of the ligand. The different orientations adopted
causes larger R-values than normally observed for other
structures. When possible, the hydrogen atoms were fixed
in positions of ideal geometry and refined within the
XSHELL software. These idealized hydrogen atoms had
their isotropic temperature factors fixed at 1.2 or 1.5 times
the equivalent isotropic U of the C atoms for which they
were bonded. The final refinement of each compound in-
cluded anisotropic thermal parameters on all non-hydro-
gen atoms. Additional information concerning the data
collection and final structural solutions of 2–11 can be
found in the supplemental information. Data collection
parameters for 2–11 are given in Table 1. Descriptions
of the problems encountered for the less well behaved
structure solutions are described below.
4.3.1. General X-ray crystal structure information for
alkoxide ligands of 2–11
Final crystal solutions that contain alkoxides are
notorious for positional disorder behavior problems.
The carbon atoms of the alkoxide ligands (OPri or
ONep) of these compounds show significant transla-
tional disorder. Attempts were made to model this disor-
der by distributing the positions of the C over several
sites. These attempts proved fruitless as the disorder
does not display clearly defined positions for the partial
occupied sites. Hence, these attempts were abandoned in
favor of one large thermal ellipsoid for many of the OR
ligands in these structures. Specific information can be
found in the Supplemental information.
4.3.2. Specific X-ray crystal structure information for 5
A coordinated solvent molecule of (H-OPri) was
observed in the lattice. Hydrogen bond lengths for the
˚
solvent were H(7)–O(11) = 1.742 A and H(11)–O(1) =
˚
1.827 A. Significant positional disorder was observed
for the methyl groups on the OBc ligand. Two possible
orientations of the methyl groups were refined resulting
in C(12)/C(120), C(13)/C(130), and C(14)/C(140) methyl
pairs each with 0.5 site occupancy.
4.3.3. Specific X-ray crystal structure information for 6
Disorder of a terminal methyl group of the ONc led
to expected bond deviations and large atomic displace-
ment parameters due to fractional occupancy as well
as ill-defined site location resulting from the ligand dis-
order (the other fraction of the site is found with the C11
terminal methyl).
4.3.4. Specific X-ray crystal structure information for 7
Two disordered hexane solvent molecules were lo-
cated within the cell and were modeled using the PLA-
TON/SQUEEZE program (ver. 01-11-99). The SQUEEZE

T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
1737
program located the solvent centers at (0.037, 3/3, 0.402)
toluene. The solution was then transferred to a syringe,
removed from the glovebox and the precursor solution
deposited onto a platinized silicon substrate. The wafer
was spun at 3000 rpm for 30 s and was then transferred
to a hot plate at 300 ꢁC to initiate crystallization. The
wafer was removed from the heat after no more color
changes were observed (ꢁ30 s), the wafer was removed,
allowed to cool, and then another layer added. This was
repeated so that 3 layers were deposited. The films were
then fired in a furnace at 550 ꢁC for one hour under
ambient atmosphere. Subsequently, the films were ana-
lyzed using standard ellipsometry techniques at three
different times during the film making process, (i) pre-
heat, (ii) post hot plate, and (iii) post fire.
and (ꢀ0.030, 1/4, 0.598), with a potential volume of
3
˚
744.2 A and electron count of 90 electrons/cell, consis-
tent with two hexane molecules.
4.3.5. Specific X-ray crystal structure information for 9
One of the terminal methyl carbon attached to a
translationally disordered OPc acid group occupies mul-
tiple site locations which leads to spatially disordered
C(55) and C(56) methyl carbons. Attempts were made
to model this disorder, however a simple disorder model
for this ligand was not possible. Partially occupied sites
tended to elongate severely and often became non-posi-
tive definite. The modeling of the disorder was abandon
in favor of the individual C(55) and C(56) carbon sites
with abnormally large thermal ellipsoids, with the
understanding that these ellipsoids represent the large
variability of methyl carbon atoms from molecule to
molecule. C(16) possesses a high Ueq in comparison to
other atoms because it is located on a terminal ONep li-
gand which has rotational disorder. The apparent rota-
tional disorder displayed on this ligand does not affect
all the carbons. In fact the rotation appears to pivot
about the central C(17) carbon which allows this carbon
to be relatively well defined; all the associated carbon
atoms bond to this atom, however tend to displace
due to the rotation. O(7) is also bound to a translation-
ally disordered yielding large thermal ellipsoids for the
associated carbons while the O7 which is more rigidly
bound to the Zr(5) atom remains relatively fixed.
Acknowledgments
Sandia is a multiprogram laboratory operated by
Sandia Corporation, a Lockheed Martin Company,
for the United States Department of Energyꢀs National
Nuclear Security Administration under contract DE-
AC04-94AL85000. The authors would also like to thank
the office of Basic Energy Sciences of the Department of
Energy for their support of this work and the technical
assistance of Dr. Todd M. Alam (SNL), Dr. Scott D.
Bunge (Kent St.), Mr. Kelly P. Peters, and Mr. Jesus
J. Gallegos, III.
Appendix A. Supplementary data
4.3.6. Specific X-ray crystal structure information for 11
One ONc ligand (O10, O11) showed rotational disor-
dering of carbon atoms and was modeled by split (0.5
occupancy) methyl C sites C36/C360 and C38/C380.
Due to the disorder, hydrogen atoms were not refined
on these split sites, however 12 hydrogens were added
to the contents of the unit cell during the final refine-
ment to properly calculate the crystal data. An H-ONep
molecule was located in the unit cell. This molecule is
centered at (0.288, 0.307, 0.370). A disordered toluene
solvent molecule was observed in the unit cell. The tol-
uene was centered at (0, 1/2, 0). Attempts to model
the disordered toluene failed. The solvent electron den-
sity was modeled using PLATON/SQUEEZE (ver. 01-11-99),
Crystallographic data for the structural analyses of
2–11 have been deposited with Cambridge Crystallo-
graphic Data Center, CCDC Nos. 272310–272319. Cop-
ies of this information may be obtained free of charge
from The Director, CCDC, 12 Union Road, Cambridge
CB2 1EZ, UK (fax: +44 1233 336 033; e-mail: deposit@
ccdc.cam.ac.uk or http://www.ccdc.cam.ac.uk). Supple-
mentary data associated with this article can be found,
in the online version at doi:10.1016/j.poly. 2005.05.005.
References
3
˚
which located a potential solvent volume of 281.4 A
and an electron count of 46 electrons/cell consistent with
one toluene molecule.
[1] H. Yoshida, M.G. Chaskar, Y. Kato, T. Hattori, J. Photochem.
Photobiol. A – Chem. 160 (2003) 47.
[2] T. Kimura, Studies Surf. Sci. Cat. 145 (2003) 485.
[3] R.D. Reeleder, B.B. Capell, L.D. Tomlinson, W.J. Hickey, Can.
J. Plant Pathol. 25 (2003) 182.
4.4. Thin film production
[4] H.K. Yueh, B. Cox, J. Nucl. Mater. 323 (2003) 57.
[5] C. Lomenech, E. Simoni, R. Drot, J.J. Ehrhardt, J. Mielczarski,
J. Colloid Interf. Sci. 261 (2003) 221.
[6] S. Pikus, E. Olszewska, S. Kalinowska, J. Appl. Crystallogr. 36
(2003) 749.
[7] J. Ciosek, W. Paszkowicz, P. Pankowski, J. Firak, U. Stanislawek,
R. Ostrowski, Proc. Soc. Phot-Optical Instrum. Eng. (SPIE) 5230
(2003) 34.
Thin films of all new compounds were generated
using standard sol–gel spin cast deposition methods
[22,38]. The general preparatory route involved generat-
ing under an argon atmosphere an ꢁ0.40 M precursor
solution with the appropriate compound dissolved in

1738
T.J. Boyle et al. / Polyhedron 24 (2005) 1727–1738
[8] N. Rana, P. Raghu, E. Shero, F. Shadman, Appl. Surf. Sci. 205
(2003) 160.
[23] D.N. Williams, U. Piarulli, C. Floriani, A. Chiesi-Villa, C. Fissoli,
J. Chem. Soc., Dalton Trans. (1994) 1243.
[9] P. Raghu, N. Rana, C. Yim, E. Shero, F. Shadman, J.
Electrochem. Soc. 150 (2003) F186.
[24] G. Kickelbick, U. Schubert, J. Chem. Soc., Dalton Trans. (1999)
1301.
[10] Z.W. Zhao, B.K. Tay, G.Q. Yu, S.P. Lau, J. Phys. Condens.
Matter 15 (2003) 7707.
[25] G. Kickelbick, P. Wiede, U. Schubert, Inorg. Chim. Acta 284
(1999) 1.
[11] C.D. Chandler, C. Roger, M.J. Hampden-Smith, Chem. Rev. 93
(1993) 1205.
[26] B. Moraru, S. Gross, G. Kickelbick, G. Rimmel, U. Shubert,
Monatsh. Chem. 132 (2001) 993.
[12] L.G. Hubert-Pfalzgraf, New. J. Chem. 11 (1987) 663.
[13] N.Y. Turova, e.P. Turevskaya, V.G. Kessler, M.L. Yanovskay,
The Chemistry of Metal Alkoxides, Kluwer Academic Publishers,
Boston, 2002.
[27] L. Ma, D.A. Payne, Chem. Mater. 6 (1994) 875.
[28] G. Kickelbick, U. Schubert, Chem. Ber. 130 (1997) 473.
[29] M.Y. Reza, H. Matsushima, M. Koikawa, M. Nakashima, T.
Tokii, Polyhedron 18 (1999) 787.
[14] D.C. Bradley, Chem. Rev. 89 (1989) 1317.
[15] D.C. Bradley, R.C. Mehrotra, A.P. Rothwell, A. Singh, Alkoxo
and Aryloxo Derivatives of Metals, Academic Press, San Diego,
2001.
[30] G. Trimmel, S. Gross, G. Kickelbick, U. Schubert, Appl.
Organomet. Chem. 15 (2001) 401.
[31] M. Veith, C. Mathur, S. Mathur, V. Huch, Organometallics 16
(1997) 1292.
[16] D.C. Bradley, R.C. Mehrotra, D.P. Gaul, Metal Alkoxides,
Academic Press, New York, 1978.
[32] W.J. Evans, G.W. Nyce, M.A. Greci, J.W. Ziller, Inorg. Chem.
40 (2001) 6725.
[17] V.W. Day, W.G. Klemperer, M.M. Pafford, Inorg. Chem. 40
(2001) 5738.
[33] W.J. Evans, M.A. Johnston, J.W. Ziller, Inorg. Chem. 39 (2000)
3421.
[18] B.A. Vaartstra, J.C. Huffman, P.S. Gradeff, L.G. Hubert-Pfalzg-
raf, J.-C. Daran, S. Parraud, K. Yunlu, K.G. Caulton, Inorg.
Chem. 29 (1990) 3126.
[19] T. Boyle, R. Schwartz, R. Doedens, J. Ziller, Inorg. Chem. 34
(1995) 1110.
[20] T.J. Boyle, J.J.I. Gallegos, D.M. Pedrotty, E.R. Mechenbeir, B.
Scott, Coord. Chem. 47 (1999) 155.
[21] T.J. Boyle, T.A. Alam, C.J. Tafoya, B.L. Scott, Inorg. Chem. 37
(1998) 5588.
[34] W.J. Evans, M.A. Greci, M.A. Johnston, Chem. A – Eur. J. 5
(1999) 3482.
[35] W.J. Evans, M.A. Greci, M.A. Ansari, J. Chem Soc., Dalton
Trans. (1997) 4503.
[36] T.J. Boyle, T.M. Alam, E.R. Mechenbeir, B. Scott, J.W. Ziller,
Inorg. Chem. 36 (1997) 3293.
[37] 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.
[22] T.J. Boyle, R.P. Tyner, T.A. Alam, B.L. Scott, J.W. Ziller, B.G.
Potter, J. Am. Chem. Soc. 121 (1999) 12104.
[38] T.J. Boyle, US Patent No. 5858451, 1991.