Mononuclear, five-coordinate lanthanide amido and aryloxide complexes bearing tetradentate (N2O2) Schiff bases

Article abstract of DOI:10.1021/ic040006f

Two monomeric, five-coordinate lanthanide complexes, [bis-5,5′-(1,3- propanediyldiimino)-2,2-dimethyl-4-hexene-3-onato]samarium[2,6-bis(tert-butyl) -4-methylphenoxide] and [bis-5,5′-(1,3-propanediyldiimino)-2,2-dimethyl-4- hexene-3-onato]erbium[2,6-bis(tert-butyl)-4-methylphenoxide], were isolated from the reactions of 2,6-bis(tert-butyl)-4-methylphenol with [bis-5,5′-(1,3- propanediyldiimino)-2,2-dimethyl-4-hexene-3-onato]lanthanide[bis(trimethylsilyl) amido] (lanthanide = Er³⁺ and Sm³⁺). The purified phenoxides were recovered in excellent yields and analytical purity, and the reactions proceeded cleanly without Schiff-base degradation or cluster formation. Analogously, [bis-3,3′-(1,3-propanediyldiimino)-1-phenyl-2- butene-1-onato]erbium[bis(trimethylsilyl)amido] was also directly converted to [bis-3,3′-(1,3-propanediyldiimino)-1-phenyl-2-butene-1-onato]erbium[2, 6-bis(tert-butyl)-4-methylphenoxide]; however, a less sterically demanding alcohol (i.e., ethanol) yielded a neutral trinuclear oxo alkoxide species with each dianionic Schiff base asymmetrically bridging through μ-oxo interactions. In this polynuclear cluster, each symmetry-related, seven-coordinate erbium(III) ion exhibits monocapped trigonal prismatic geometry, which assembles by sharing triangular capped faces. Single-crystal X-ray diffraction revealed square-pyramidal metal coordination in each five-coordinate lanthanide ion with varied S₄ ruffling of the square base donor atoms and the six-membered propylene diamine chelate ring adopting the boat conformation. To contrast the effect of subtle ligand changes, we also report the synthesis and characterization of [bis-5,5′-(2,2-dimethyl-1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-3- onato]samarium[bis(trimethylsilyl)amido], having gem-dimethyl substituents appended to the propylene bridge central carbon. The six-membered diamine chelate ring in this compound adopts the chair conformation without metal-hydrocarbon interaction. Also presented are qualitative activity observations and polymerization data for the polymerization of rac-lactide and ε-caprolactone using the five-coordinate lanthanide amidos and phenoxides.

Full text of DOI:10.1021/ic040006f

Inorg. Chem. 2004, 43, 6203−6214
Mononuclear, Five-Coordinate Lanthanide Amido and Aryloxide
Complexes Bearing Tetradentate (N2O2) Schiff Bases
Steven A. Schuetz,† Carter M. Silvernail,† Christopher D. Incarvito,‡ Arnold L. Rheingold,‡
Joanna L. Clark,† Victor W. Day,† and John A. Belot*,†
Department of Chemistry and the Center for Materials Research and Analysis, UniVersity of
Nebraska, Lincoln, Nebraska 68588, and Department of Chemistry and Biochemistry,
UniVersity of Delaware, Newark, Delaware 19716
Received January 21, 2004
Two monomeric, five-coordinate lanthanide complexes, [bis-5,5
onato]samarium[2,6-bis(tert-butyl)-4-methylphenoxide] and [bis-5,5
3-onato]erbium[2,6-bis(tert-butyl)-4-methylphenoxide], were isolated from the reactions of 2,6-bis(tert-butyl)-4-
′
-(1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-3-
′
-(1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-
methylphenol with [bis-5,5
′
-(1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-3-onato]lanthanide[bis(trimethylsilyl)amido]
+
(lanthanide
)
Er3 and Sm3+). The purified phenoxides were recovered in excellent yields and analytical purity,
and the reactions proceeded cleanly without Schiff-base degradation or cluster formation. Analogously, [bis-3,3
(1,3-propanediyldiimino)-1-phenyl-2-butene-1-onato]erbium[bis(trimethylsilyl)amido] was also directly converted to
[bis-3,3 -(1,3-propanediyldiimino)-1-phenyl-2-butene-1-onato]erbium[2,6-bis(tert-butyl)-4-methylphenoxide]; however,
a less sterically demanding alcohol (i.e., ethanol) yielded a neutral trinuclear oxo alkoxide species with each dianionic
Schiff base asymmetrically bridging through -oxo interactions. In this polynuclear cluster, each symmetry-related,
′-
′
µ
seven-coordinate erbium(III) ion exhibits monocapped trigonal prismatic geometry, which assembles by sharing
triangular capped faces. Single-crystal X-ray diffraction revealed square-pyramidal metal coordination in each five-
coordinate lanthanide ion with varied S4 ruffling of the “square base” donor atoms and the six-membered propylene
diamine chelate ring adopting the boat conformation. To contrast the effect of subtle ligand changes, we also
report the synthesis and characterization of [bis-5,5
3-onato]samarium[bis(trimethylsilyl)amido], having gem-dimethyl substituents appended to the propylene bridge central
carbon. The six-membered diamine chelate ring in this compound adopts the chair conformation without metal
hydrocarbon interaction. Also presented are qualitative activity observations and polymerization data for the
′-(2,2-dimethyl-1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-
−
polymerization of rac-lactide and
ꢀ
-caprolactone using the five-coordinate lanthanide amidos and phenoxides.
Introduction
anhydrous, hard donor-atom (i.e., N or O) frameworks have
been the subject of recent investigations; relevant examples
A common theme in contemporary f-element chemistry
is identifying alternative ligand scaffolds beyond cyclopen-
tadienyl-based coordination environments that maintain the
low coordination numbers and catalytic activity of organo-
metallic Ln3+ complexes.1a-e Toward this goal, various
include heteroleptic guanidinates,2a,b calix[4]arenes,3 por-
phyrins,4 â-diketiminato,5 and tetradentate Schiff bases.6a-k
Of these potential organic scaffolds, the latter offer the
distinct advantages of ease of synthesis, low cost, rapid
tailorability at both peripheral and prochiral backbone sites,
* Author to whom correspondence should be addressed. E-mail: jbelot2@
unl.edu.
(2) (a) Giesbrecht, G. R.; Whitener, G. D.; Arnold, J. J. Chem. Soc., Dalton
Trans. 2001, 923. (b) Noss, H.; Oberthu¨r, M.; Fischer, C.; Kretschmer,
W. P.; Kempe, R. Eur. J. Inorg. Chem. 1999, 2283.
(3) Estler, F.; Herdtweck, E.; Anwander, R. J. Chem. Soc., Dalton Trans.
2002, 3088.
(4) Foley, T. J.; Abboud, K. A.; Boncella, J. M. Inorg. Chem. 2002, 41,
1704.
† University of NebraskasLincoln.
‡ University of Delaware. Present address: Departments of Chemistry
and Biochemistry, University of California-San Diego, La Jolla, CA 92093.
(1) (a) Dehnicke, K.; Greiner, A. Angew. Chem., Int. Ed. 2003, 42, 1340.
(b) Piers, W. E.; Emslie, J. J. H. Coord. Chem. ReV. 2002, 233-234,
131. (c) Aspinall, H. C. Chem. ReV. 2002, 102, 1807. (d) Roesky, P.
W. Chem. Soc. ReV. 2000, 29, 335. (e) Evans, W. J. New J. Chem.
1995, 19. 525.
(5) Neculai, A. M.; Roesky, H. W.; Neculai, D.; Magull, J. Organome-
tallics 2001, 20, 5501.
10.1021/ic040006f CCC: $27.50
Published on Web 09/04/2004
© 2004 American Chemical Society
Inorganic Chemistry, Vol. 43, No. 20, 2004 6203

Schuetz et al.
Figure 1. General diffraction-derived structures of mononuclear amino complexes containing methylene (A) and gem-dimethyl (B) substituents along the
pseudomolecular mirror plane.
and ample literature precedent.7 However, isolating low
coordination number, anhydrous, f-element complexes bear-
ing these ligands is fraught with obstacles and undoubtedly
contributes to the paucity of literature reports.1b,c For
example, anhydrous Salen-based lanthanide complexes often
utilize high coordination numbers (>6) via dinuclear Ln2L3
formation (L ) tetradentate Salen), whose formation is
believed to be enhanced by π stacking.8a-b Lewis base
solvent stabilization of the coordinately unsaturated metal
center also presents higher coordination number ions6a but
stabilizes mononuclear Salen complexes containing group
III elements.6d,9a,b Regarding the latter, Evans6d reported a
variety of solvated Y3+ Salen-based compounds that are
particularly germane to this study.
The importance of synthesizing mononuclear lanthanide
complexes and understanding the conditions required to
promote their formation is that the complexes can be utilized
as catalysts for small molecules and the polymerization of
polar molecules (vide infra). With the new, readily available,
low-cost, easily tailored lanthanide Schiff-base catalysts
described herein, there is seemingly unlimited potential to
tailor ligands at multiple sites as well as metal apical
substituents to enable small-molecule transformations ex-
ploiting f-element attributes. We previously reported the
synthesis and characterization of mononuclear Sm3+ (1)6e and
Er3+ (2) amido complexes bearing “saturated” Schiff bases
(Figure 1A and B, respectively) as well as the stoichiometric
hydrolysis chemistry of the mononuclear erbium amido, 2.6c
Inspired by the subtle structural differences within these
molecules and the significant amido lability, we began to
evaluate alternative substitution patterns on the Schiff-base
(SB) periphery as well as the direct replacement of the bulky
bis(trimethylsilyl)amido group using various Brønsted acids.
Specifically, we sought to yield more accessible rare-earth
ions for a variety of substrates by decreasing ketoiminato
(KI) frontal steric bulk through tBu group replacement. We
also postulated increased axial-ligand nucleophilicity toward
incoming substrates in mononuclear (KI)Ln(alkoxide) spe-
cies. Although the ability to tailor Schiff bases is the crux
of their desirability,1c,7 identifying the lower limit of steric
bulk required to recover stable mononuclear species suc-
cessfully for small charge-to-radius ratio ions remains
challenging. Furthermore, we sought to establish that the
mononuclear amido complexes would undergo ligand ex-
change reactions without ketoiminato degradation, replace-
ment, or protonolysis by alcohols,5 thereby proving them-
selves to be versatile, stable platforms for future work
involving the polymerization of cyclic ester or epoxide
monomers or small-molecule transformations.
(6) (a) Schuetz, S. A.; Day, V. W.; Rheingold, A. L.; Belot, J. A. J. Chem.
Soc., Dalton Trans. 2003, 22, 4303. (b) Yu, L.-B.; Yao, Y.-M.; Shen,
Q.; Zhang, J.; Wu, L.-X.; Ye, L. Chin. J. Chem. 2003, 21, 442. (c)
Schuetz, S. A.; Day, V. W.; Clark, J. L.; Belot, J. A. Inorg. Chem.
Commun. 2002, 5, 706. (d) Evans, W. J.; Fujimoto, C. H.; Ziller, J.
W. Polyhedron 2002, 21, 1683. (e) Schuetz, S. A.; Day, V. W.;
Sommer, R. D.; Rheingold, A. L.; Belot, J. A. Inorg. Chem. 2001,
40, 5292. (f) Evans, W. J.; Fujimoto, C. H.; Ziller, J. W. J. Chem.
Soc., Chem. Commun. 1999, 311. (g) Liu, Q.; Ding, M.; Lin, Y.; Xing,
Y. Polyhedron 1998, 13-14, 2327. (h) Liu, Q.; Ding, M. J.
Organomet. Chem. 1998, 553, 179. (i) Liu, Q.; Ding, M.; Lin, Y.;
Xing, Y. Polyhedron, 1998, 17, 555. (j) Dube, T.; Gambarotta, S.;
Yap, G. Organometallics 1998, 17, 3967. (k) Liu, Q.; Ding, M.; Lin,
Y.; Xing, Y. J. Organomet. Chem. 1997, 548, 139.
(7) Yamada, S. Coord. Chem. ReV. 1999, 190-192, 537.
(8) (a) Cai, Y. P.; Ma, H. Z.; Kang, B. S.; Su, C. Y.; Zhang, W.; Sun, J.;
Xiong, Y. L. J. Organomet. Chem. 2001, 628, 99. (b) Blech, P.;
Floriani, C.; Chiesivilla, A.; Guastini, C. J. Chem. Soc., Dalton Trans.
1990, 3557.
(9) (a) Lin, M.-H.; RajanBabu, T. V. Org. Lett. 2002, 4, 1607. (b) Runte,
O.; Priermeier, T.; Anwander, R. J. Chem. Soc., Chem. Commun. 1996,
1385.
Herein, we present the results of stoichiometric reactions
between alcohols and three unique rare-earth Schiff-base
amido complexes, [bis-5,5′-(1,3-propanediyldiimino)-2,2-di-
methyl-4-hexene-3-onato]samarium[bis(trimethylsilyl)ami-
do] (1), [bis-5,5′-(1,3-propanediyldiimino)-2,2-dimethyl-4-
hexene-3-onato]erbium[bis(trimethylsilyl)amido] (5), and
[bis-3,3′-(1,3-propanediyldiimino)-1-phenyl-2-butene-1-ona-
to]erbium[bis(trimethylsilyl)amido] (6). To illustrate the
effect of subtle alkyl substitutions on the ligand backbone
further, the synthesis and characterization of [bis-5,5′-(2,2-
dimethyl-1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-3-
onato]samarium[bis(trimethylsilyl)amido] (4) is included. In
4, the central carbon (C(6)) of the propylene bridge supports
6204 Inorganic Chemistry, Vol. 43, No. 20, 2004

Lanthanide Amido and Aryloxide Complexes
Calcd for C23H26N2O2: C, 76.21; H, 7.23; N, 7.73. Found: C, 76.59;
H, 7.10; N, 7.78.
gem-dimethyls, which significantly changes the six-mem-
bered diamino chelate-ring geometry. Complexes 1 and 5
differ only in their rare-earth elements, whereas 6 has phenyl
substituents replacing the bulky, frontal tBu groups of 5. This
modification not only makes the rare-earth ion more acces-
sible to substrate approach in both the derived amino and
alkoxide complexes but also creates more freedom for axial-
ligand motion. These syntheses explicitly demonstrate that
this general ligand scaffold tolerates both various peripheral
substituents and changing Ln3+ radii in affording discrete,
mononuclear complexes. The isolated alkoxides also present
mononuclear complexes when sterically demanding aryl
alcohols occupy the apical coordination site; however, when
the alcohol bulk is drastically reduced (i.e., EtOH) the
formation of a highly symmetric, neutral trinuclear species
is observed. All of these complexes are fully characterized
using a variety of spectroscopic techniques including el-
emental analyses and single-crystal X-ray diffraction and are
amenable to large-scale, reproducible preparations.
Synthesis of [Bis-5,5′-(2,2-dimethyl-1,3-propanediyldiimino)-
2,2-dimethyl-4-hexene-3-onato]samarium[bis(trimethylsilyl)-
amido] (4). A 14/20, 100-mL Schenk flask is charged with 0.823
g (1.30 mmol) of tris(bis(trimethylsilyl)amido)samarium (glovebox),
sealed with a septum, interfaced to a Schlenk line, and charged
with 40 mL of C7H16. In a separate 14/20, 100-mL round-bottom
flask, 0.457 g (1.30 mmol) of bis-5,5′-(2,2-dimethyl-1,3-pro-
panediyldiimino)-2,2-dimethyl-4-hexene-3-one is dissolved in 40
mL of C7H16. The ligand solution is then added to the samarium
flask via syringe in one aliquot. Under flowing N2, the septum is
replaced with a reflux condenser, and the reaction is heated to reflux
for 12 h. The homogeneous, dull-yellow mother liquor is then
concentrated in vacuo (∼10 mL) and cooled to -10 °C. Pale-yellow
crystals are harvested after 4 days. Yield: 67.1% (0.561 g); mp:
1
152-154 °C; H NMR (δ, C6D6): -4.14, -2.30, -2.02, -0.79,
0.08, 0.15, 0.60, 0.66, 0.76, 0.86, 1.05, 1.11, 1.27, 1.41, 1.55, 1.87,
2.12, 4.13, 4.59, 6.49, 6.88, 7.73, 10.61; 13C NMR (δ, C6D6): 5.9,
18.9, 20.5, 28.1, 29.2, 29.6, 54.3, 99.9, 174.9, 195.9. Anal. Calcd
for C27H54N3O2Si2Sm: C, 49.19; H, 8.26. Found: C, 49.21; H,
8.15.
Experimental Section
Synthesis of [Bis-5,5′-(1,3-propanediyldiimino)-2,2-dimethyl-
4-hexene-3-onato]erbium[bis(trimethylsilyl)amido] (5). To a 250-
mL Schlenk flask containing 1.81 g (2.79 mmol) of tris(bis-
[trimethylsilyl]amido)erbium dissolved in 100 mL of C7H16, we
added 0.900 g (2.79 mmol) of bis-5,5′-(1,3-propanediyldiimino)-
2,2-dimethyl-4-hexene-3-one into 50 mL of C7H16 via a syringe.
The solution was stirred and heated to reflux overnight, the organics
were removed in vacuo, and the crude product was recrystallized
(24 h) from 10 mL of C5H12 at -10 °C as pale-pink crystals.
Materials and Methods. All reactions were carried out under
an atmosphere of dry dintrogen (passage over two 1.5-m glass
columns containing Drierite and molecular sieves) using standard
Schlenk and glovebox techniques. The synthesis of [tris(bis-
(trimethylsilyl)amido]lanthanides and bis-5,5′-(1,3-propanediyldi-
imino)-2,2-dimethyl-4-hexene-3-one were accomplished using lit-
erature preparations.6e 2,6-Di(tert-butyl)-4-methyl phenol (Aldrich)
was doubly sublimed (∼10-4-10-5 Torr) before use and stored in
the glovebox, whereas 1-phenyl-1,3-butanedione (Aldrich) was used
as received. Heptane (C7H16, EM Science) and pentane (C5H12,
Fisher) were dried and distilled from CaH2. Toluene (C7H8) and
1
Yield: 82% (1.48 g); mp: 153.8-154.9 °C; H NMR (δ, C6D6):
0.05 (s, 18H), 1.00 (m, 2H), 1.34 (b, 6H), 4.90 (m, 4H), 6.74 (b,
2H), 10.62 (b, 2H); 13C NMR (δ, C6D6): 2.6, 6.1, 19.2, 20.8, 29.4,
29.8, 37.2, 54.6, 100.1, 175.2, 196.2; Anal. Calcd for C25H50N3-
O2Si2Er: C, 46.33; H, 7.78. Found C, 46.55; H, 7.61.
ethanol (EtOH) were dried and distilled from sodium. H and 13C
1
NMR were recorded on a Bruker DRX Avance 500-MHz spec-
trometer (1H frequency) or a DRX Avance 400-MHz spectrometer
(1H frequency) using J. Young NMR tubes into which dry
deuterated solvents were vacuum transferred from storage bulbs
containing a Na/K alloy. Elemental analyses were performed by
Midwest Microlabs (Indianapolis, IN), and the melting points
(uncorrected) were collected on a modified Mel-Temp II apparatus
providing digital thermocouple readouts.
Synthesis of [Bis-3,3′-(1,3-propanediyldiimino)-1-phenyl-2-
butene-1-onato]erbium[bis(trimethylsilyl)amido] (6). In a 100-
mL Schlenk flask, 1.119 g (1.726 mmol) of tris[bis(trimethylsilyl)-
amido]erbium was dissolved in 30 mL of C7H8. To this solution,
we added 0.6256 g (1.726 mmol) of 3 dissolved in warm C7H8.
The reaction was heated to reflux overnight, cooled to room
temperature, and concentrated in vacuo, and the title compound
was recrystallized from the mother liquor after 12 h at -18 °C as
pink crystals. Yield: 60.3% (0.7043 g); mp: 140-142 °C (dec);
1H NMR (δ, C6D6): -42.16, -28.59, 0.01, 2.09, 4.53, 7.02, 10.28,
13.73, 27.93; 13C NMR (δ, C6D6): -63.61, -43.22, -32.63, 2.34,
5.80, 21.29, 22.00, 79.46, 106.65, 125.71, 135.58; Anal. Calcd for
C29H42N3O2Si2Er ‚1/2C7H8: C, 53.17; H, 6.32. Found C, 53.11; H,
6.22 (the presence of toluene was confirmed by X-ray diffraction).
Representative Aryloxide Synthesis of [Bis-5,5′-(1,3-pro-
panediyldiimino)-2,2-dimethyl-4-hexene-3-onato]lanthanide[2,6-
bis(tert-butyl)-4-methylphenoxide]. In a 100-mL Schlenk flask
0.47 mmol [bis-5,5′-(1,3-propanediyldiimino)-2,2-dimethyl-4-hex-
ene-3-onato]lanthanide[bis(trimethylsilyl)amido] was dissolved in
40 mL of C5H12, and 0.11 g (0.47 mmol) of 2,6-di(tert-butyl)-4-
methyl phenol in 40 mL of C5H12 was added via a syringe. The
solution was stirred and heated to reflux overnight, the organics
were removed in vacuo, and the crude products were recrystallized
(24 h) from ∼5 mL of C7H8 at -18 °C.
Synthesis of Bis-3,3′-(1,3-propanediyldiimino)-1-phenyl-2-
butene-1-one (3).10 In a 500-mL round-bottom flask, 20 g (0.12
mol) of 1-phenyl-1,3-butanedione was dissolved in 300 mL of
EtOH. An addition funnel was charged with 4.6 g (0.062 mol) of
1,3-diaminopropane in 75 mL of EtOH, and this solution was added
dropwise (2 drops/s) to completion. The reaction was heated to
reflux overnight, and then cooled and poured into 250 mL of H2O,
affording a white precipitate. This precipitate was isolated by simple
vacuum filtration on a glass frit funnel (M porosity) and recrystal-
lized from 200 mL of dry, boiling C7H16 as colorless needles. The
collected crystals were then dissolved in 200 mL of C6H6 and further
dried by continuous azeotropic distillation until a constant boiling
point was observed. The resulting ligand was again dried in vacuo
(10-4-10-5 Torr at 40 °C) for 24 h. Yield: 41% (9.2 g); mp: 97-
100 °C; 1H NMR (δ, CDCl3): 2.03 (m, 2H), 2.09 (s, 6H), 3.50 (m,
2H), 5.71 (s, 2H), 7.39 (b, 1H), 7.40 (b, 1H), 7.41 (b, 1H), 7.41 (b,
1H), 7.42 (b, 1H), 7.84 (b, 1H), 7.85 (b, 1H), 7.86 (b, 1H), 7.86
(b, 1H), 7.87 (b, 1H), 11.54 (b, 2H); 13C NMR (δ, CDCl3): 19.6,
30.3, 40.2, 92.7, 127.1, 128.4, 130.5, 140.5, 165.2, 188.4; Anal.
[Bis-5,5′-(1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-3-
onato]erbium[2,6-bis(tert-butyl)-4-methylphenoxide] (7 and Sol-
vent Polymorph 9). Yield: 82% (0.28 g, pale pink); mp: 158.0-
Inorganic Chemistry, Vol. 43, No. 20, 2004 6205

Schuetz et al.
Table 1. Crystallographic Data for Compounds 4-11a
4
5
6
7
formula
fw
C27H54N3O2Si2Sm
659.26
C25H50N3O2Si2Er
648.12
C32.50H46N3O2Si2Er
734.16
C41H63N2O3Er
799.19
cryst syst,
space groupb
a (Å)
b (Å)
c (Å)
monoclinic,
P21/c (No. 14)
12.319(3)
16.058(4)
17.752(4)
90.000
109.132(3)
90.000
3317.8(13)
4, 1.320
monoclinic,
P21/c (No. 14)
10.563(1)
25.738(1)
12.000(1)
90.000
107.127(1)
90.000
3117.7(3)
4, 1.381
triclinic,
monoclinic,
P21/n (No. 14f)
11.831(2)
21.780(4)
15.584(3)
90.000
92.397(4)
90.000
4012.2(13)
4, 1.323
P1h (No. 2)
14.245(1)
15.547(1)
17.105(1)
105.201(1)
102.598(1)
100.237(1)
3454.7(3)
4, 1.412
R (deg)
â (deg)
γ (deg)
V (Å3)
Z, Fcalcd (Mg/m3)
µ (Mo KR) (mm-1
cryst size (mm3)
data completeness
)
1.867
2.792
2.529
2.128
0.21 × 0.10 × 0.09
0.21 × 0.20 × 0.12
0.10 × 0.09 × 0.08
0.10 × 0.08 × 0.06
95.8%
99.8%
99.7%
99.8%
(max θMo KR
)
(28.32)
(28.35)
(28.30)
empirical
0.914/1.000
1.223/-0.809
173(2)
(28.33)
absorption correction
min/max transmission
largest diff. peak/hole (e-/A3)
temp (K)
empirical
0.604/0.415
2.068/-1.060
100(2)
empirical
0.896/1.000
0.831/-0.566
173(2)
empirical
0.775/1.000
2.399/-1.549
173(2)
no. of reflns (collected/unique)
38 927/7898
0.034(6555)
0.076(7898)
0.997
33 544/7757
0.024(6535)
0.048(7757)
0.948
50 028/17 139
0.0321 (12571)
0.0564(17 139)
0.830
46 618/9990
0.052(7729)
0.100(9990)
0.990
R1c (I > 2σI)
wR2d (unique reflns)
GOF indicatore
8
9
10
11
formula
fw
C41H63N2O3Sm
782.28
C41H63N2O3Er
753.13
C48.50H59N2O3Er
885.24
C92H101N6O8Er3
1920.57
cryst syst,
space groupb
a (Å)
b (Å)
c (Å)
monoclinic,
P21/n (No. 14f)
11.715(2)
22.216(3)
15.548(2)
90.000
92.469(2)
90.000
4042.9(9)
4, 1.285
monoclinic,
P21/n (No. 14f)
13.600(1)
20.226(1)
14.231(1)
90.000
106.035(1)
90.000
3762.1(3)
4, 1.330
triclinic,
monoclinic,
P2(1)/c (No. 14)
21.573(2)
16.529(2)
23.422(3)
90.000
93.678(2)
90.000
8334.7(16)
4, 1.531
P1h (No. 2)
13.602(1)
13.700(1)
13.708(1)
78.515(5)
60.973(3)
78.482(3)
2173.0(2)
2, 1.353
R (deg)
â (deg)
γ (deg)
V (Å3)
Z, Fcalcd (Mg/m3)
µ (Mo KR) (mm-1
cryst size (mm3)
data completeness
(max θMo KR)
)
1.489
2.265
1.972
3.052
0.30 × 0.20 × 0.20
0.16 × 0.10 × 0.09
0.10 × 0.07 × 0.05
0.50 × 0.32 × 0.32
93.6%
99.9%
99.3%
99.9%
(28.30)
(28.30)
(28.31)
(28.34)
empirical
0.704/1.000
1.797/-0.642
173(2)
absorption correction
min/max transmission
largest diff peak/hole (e-/A3)
temp (K)
SADABS
0.759/1.000
1.095/-0.841
150(2)
empirical
0.850/1.000
1.054/-0.331
173(2)
empirical
0.922/1.000
1.837/-0.654
173(2)
no. of reflns (collected/unique)
25 140/9400
0.042(7206)
0.095(9400)
1.036
40 286/9322
0.0253(7395)
0.054(9322)
0.891
25 221/10 729
0.051(8441)
0.110(10 729)
0.977
115 421/20 772
0.032(17 135)
0.091(20 772)
1.062
R1c (I > 2σI)
wR2d (unique reflections)
GOF indicatore
a G. M. Sheldrick, SHELXTL, version 5.1; Bruker-AXS: Madison, WI, 1999. b Henry, N. F. M.; Lonsdale, K., Eds. International Tables for X-ray
2
Crystallography, Vol. 1: Symmetry Groups; Kyntoch Press, U.K., 1969. c R1 ) ∑ | Fo| - | Fc | |/∑ | Fo |. d R2 ) [(∑w(| Fo| - | Fc |)2/∑w(| Fo| )]1/2.e GOF:
5
[(∑w(| Fo| - | Fc |)2/(Nobs - Nparam)]1/2
.
f P21/n is an alternate setting for P21/c - C2h (No. 14).
160.5 °C (dec); 1H NMR (δ, C6D6): -53.28 (b, H), -13.46 (b, H),
-6.20 (b, H), -2.43 (b, H), 1.39 (s, H), 2.22 (s, H), 2.29 (b, H),
4.77 (b, H), 7.12 (s, H), 7.20 (s, H), 8.86 (b, H); 13C NMR (δ,
C6D6): -55.7, -42.9, -21.1, -16.8, 13.4, 21.4, 29.9, 30.3, 40.7,
73.4, 95.6, 101.0, 125.7, 129.3, 137.9; Anal. Calcd for C34H55N2O3-
Er‚1/3C7H8: C, 59.15; H, 7.88. Found C: 59.30; H, 7.82 (the
presence of toluene was confirmed by X-ray diffraction).
197.44; Anal. Calcd for C34H55N2O3Sm: C, 59.17; H, 8.03. Found
C, 59.29; H, 7.98.
[Bis-3,3′-(1,3-propanediyldiimino)-1-phenyl-2-butene-1-onato]-
erbium[2,6-bis(tert-butyl)-4-methylphenoxide] (10). Yield: 71%
(0.23 g, pale pink); mp: 161.0-163.0 °C (dec); 1H NMR (δ, C6D6):
-49.82 (b), -14.51 (b), -6.99 (b), -5.70 (b), 1.18 (b), 2.02 (b),
2.16 (b), 4.44 (b), 6.90 (b), 11.91 (b), 16.70 (b), 37.57 (b); 13C
NMR (δ, C6D6): 137.13, 135.77, 131.41, 125.64, 98.51, 92.56,
67.37, 66.64, 34.06, 30.22, 28.44, 21.28, 12.313, -17.58, -46.19,
-60.44, -145.52; Anal. Calcd for ErN2C38H47O3: C, 61.09; H,
6.34. Found: C, 60.69; H, 6.37.
[Bis-5,5′-(1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-3-
onato]samarium[2,6-bis(tert-butyl)-4-methylphenoxide] (8). Yield:
1
81% (0.26 g, pale yellow); mp: 153.5-155.0 °C (dec); H NMR
(δ, C6D6) -23.84 (b, H), -6.56 (b, H), -3.68 (b, H), -2.79 (b,
H), 0.84 (s, 2H), 0.99 (s, H), 1.38 (b, 2H), 2.09 (s, H), 2.22 (b,
2H), 2.73 (s, 3H), 3.52 (b, H), 4.78 (b, 2H), 7.05 (b, H), 8.02 (b,
H), 8.51 (b, H); 13C NMR (δ, C6D6) 20.35. 22.39, 28.89, 33.78,
36.76, 38.04, 43.40, 101.43, 124.91, 126.71, 139.18, 163.54, 178.82,
Synthesis of the Er-Oxo-Alkoxy Cluster (11). In a 100-mL
Schlenk flask, 1.0 g (1.4 mmol) of 7 was dissolved in 30 mL of
C7H8. To this solution, we added 0.065 g (1.4 mmol, 0.083 mL) of
dried EtOH dissolved in 30 mL of C7H8. This solution was then
6206 Inorganic Chemistry, Vol. 43, No. 20, 2004

Lanthanide Amido and Aryloxide Complexes
Table 2. Average Metal-Ligand Bond Lengths (Å) in the Coordination Polyhedra of 4-10
Schiff base
axial ligand
compd
type
length
type
length
type
length
2.307(2)
4
8
Sm-O
Sm-O
Sm-O
Er-O
Er-O
Er-O
Er-O
Er-O
Er-O
2.218(2,9,9,2)
2.206(2,0,0,2)
2.212(2,14,7,4)
2.138(2,3,3,2)
2.143(2,13,7,4)
2.140(3,8,8,2)
2.131(2,1,1,2)
2.144(3,18,18,2)
2.140(2,21,9,12)
Sm-N
Sm-N
Sm-N
Er-N
Er-N
Er-N
Er-N
Er-N
Er-N
2.466(2,6,6,2)
2.464(3,3,3,2)
2.465(3,7,5,4)
2.371(2,2,2,2)
2.371(3,10,6,4)
2.369(4,6,6,2)
2.366(2,8,8,2)
2.366(3,13,13,2)
2.369(3,16,7,12)
Sm-N
Sm-O
2.143(2)
average
5
6
7
9
10
average
Er-N
Er-N
Er-O
Er-O
Er-O
Er-O
2.230(2)
2.226(3,1,1,2)
2.073(3)
2.081(2)
2.071(3)
Table 3. Structurally Significant Parameters Derived from Final
Atomic Coordinates of 4-1016
stirred overnight and concentrated in vacuo, and pale-pink crystals
formed after 7 days at -18 °C. H NMR (δ, C6D6): -28.08 (b),
0.049 (s), 0.95 (b), 1.24 (b,d), 2.07 (s), 5.08 (b), 6.98 (s), 7.05 (s),
7.30 (b), 10.17 (b), 13.56 (b), 27.49 (s); Anal. Calcd for C71H77N6O8-
Er3‚C7H8: C, 53.96; H, 4.93. Found: C, 53.06; H, 5.99.
1
compd
fold 1a
28.3, 29.6 4.0, 7.4
25.7, 44.1 8.7, 16.3 1.6 0.36 0.19 0.89 2.82, 2.73
fold 2b
ax
τ
∆L ∆M
∆Me
4
5
4.1 0.08 0.05 0.83 2.84, 2.82
6 molecule 1 32.1, 44.8 3.2, 15.8 3.7 0.32 0.17 0.92 2.87, 2.80
General Homopolymerizations of rac-Lactide and ꢀ-Capro-
lactone. In a 100-mL Schlenk flask either 1.44 g of rac-lactide or
1.14 g of ꢀ-caprolactone (10.0 mmol) was dissolved in 65 mL of
C7H8 (0.154 M). While stirring, 2.00 mL of a 0.0500 M toluene
catalyst solution (6.0 mL of a 0.0167M solution for ꢀ-caprolactone
polymerization) was added in one portion, and the reaction was
stirred at either 26 or 70 °C for 30 min. The polymerization reaction
was subsequently quenched with 20 mL of iPrOH, and the polymer
was precipitated with excess pentane (∼150 mL). The resulting
colorless solid was dried in vacuo (26 °C at ∼10-4-10-5 Torr) to
a constant weight.
6 molecule 2 33.2, 38.4 2.2, 9.3
2.0 0.17 0.09 0.89 2.76, 2.82
7.6 0.02 0.01 0.86 3.15, 3.53
3.38, 3.51
7
38.1, 36.2 9.8, 9.2
8
38.7, 39.6 11.0, 10.4 4.0 0.10 0.06 0.93 3.38, 3.57
3.41, 3.77
36.5, 33.0 10.6, 5.5 3.9 0.14 0.07 0.82 3.36, 3.42
3.35, 3.35
9
10
25.7, 21.6 2.5, 6.9
1.6 0.04 0.14 0.80 3.26, 3.23
3.26, 3.63
a Defined as the dihedral angle between the square base and seven-atom
ketoiminate half. b Defined as the fold of the KI chelate ring about the Ln3+
ON vector.
X-ray Structure Determinations. Lattice constants, space
groups, final agreement factors, and other relevant crystallographic
data for 4-11 are given in Table 1. Complete hemispheres of
diffraction data were collected for each compound with a Bruker
SMART APEX CCD area detector using 0.30°-wide ω scans and
graphite-monochromated Mo KR radiation. X-rays for each study
were provided by a fine-focus-sealed X-ray tube operated at 50
kV and 40 mA. Frame data (10 s, 0.30°-wide ω scans) were
collected using the Bruker SMART software package, and final
lattice constants were determined with the Bruker SAINT software
package. The Bruker SHELXTL-NT software package was used
to solve and refine each structure. “Direct methods” was used to
solve each structure, and the resulting parameters were refined with
F2 data to convergence using counterweighted full-matrix least-
squares techniques and a structural model that incorporated aniso-
tropic thermal parameters for all ordered nonhydrogen atoms,
including most of the solvent nonhydrogen atoms. Isotropic thermal
parameters were incorporated for all hydrogen atoms and several
carbon atoms of disordered toluene solvent molecules of crystal-
lization. The highest peaks (2.40-0.83 e-/Å3) in the final difference
Fouriers for 4-11 were within 1.07 Å of a metal atom or a carbon
atom of a disordered toluene molecule of crystallization. Discussions
outlining the treatment of hydrogen atoms and solvent disorder in
structures 4-11 are given in Supporting Information.
Crystals of 4-11 have several noteworthy features. Crystals of
4 and 5 are isomorphous with previously reported Er3+ and Sm3+
complexes6c,e and although the lattice constants and space groups
for 7 and 8 would tend to indicate another isomorphous pair, this
is not the case. If crystals of 7 and 8 contained the same metal,
then they would be rigorous polymorphs. The asymmetric units of
7 and 9 contain different numbers of toluene solvent molecules of
crystallization; these two crystalline forms of solvated [O2N2C19-
H32]Er[OC6H2CH3(t-C4H9)2] are therefore solvate pseudopoly-
morphs. The two crystallographically independent [O2N2C23H24]-
Er[N(Si(t-C4H9)3)2] molecules in the asymmetric unit of 6 are related
1
by a noncrystallographic inversion center near (1/4, 0, /4) in the
unit cell, and the center of the ring for the toluene solvent molecule
of crystallization is located near (3/4, 1/2, 1/4) in the unit cell. Atoms
in the asymmetric unit of 6 are labeled to reflect this pseudosym-
metry. Average bond lengths in the coordination spheres of 4-10
are given with estimated standard deviations in Table 2. Structurally
significant parameters derived from the final atomic coordinates
of 4-10 are presented in Table 3.
Results and Discussion
The three acyclic ketoiminates used to ligate the anhy-
drous, solvent-free rare earth ions, bis-3,3′-(1,3-propanediyl-
diimino)-1-phenyl-2-butene-1-one (3), bis-5,5′-(1,3-pro-
panediyldiimino)-2,2-dimethyl-4-hexene-3-one, and bis-5,5′-
(2,2-dimethyl-1,3-propanediyldiimino)-2,2-dimethyl-4-hexene-
3-one6c were synthesized by reacting the appropriate diketone
with 1,3-propanediamine.7,10 The diketone for 3 is com-
mercially available, but the starting material for the two
others must be isolated from the NaH-initiated Claisen
condensation of pinacolone with ethyl acetate.11 Although
the chelators are inexpensively and straightforwardly syn-
thesized and isolated in high yields and analytical purity
(note: rigorous removal of adventitious H2O and EtOH must
be undertaken via azeotropes), which adds to their desir-
ability, these frameworks were not designed for convenience
but rather to accommodate large f elements exhibiting low
(10) (a) McCarthy, P. J.; Martell, A. E. Inorg. Chem. 1967, 6, 781. (b)
McCarthy, P. J.; Hovey, R. J.; Ueno, K.; Martell, A. E. J. Am. Chem.
Soc. 1955, 77, 5820. (c) Although this ligand is known, we report its
synthesis and characterization as the necessary H2O- and EtOH-free
compound.
(11) Adams, J. T.; Hauser, C. R. J. Am. Chem. Soc. 1944, 66, 1220.
Inorganic Chemistry, Vol. 43, No. 20, 2004 6207

Schuetz et al.
Scheme 1. Alcoholysis Reaction Pathways and Products Involving Mononuclear Amino Complexes 1, 5, and 6a
a The HN(TMS)2 byproduct has been omitted.
coordination numbers. Specifically, on the basis of failed
attempts to isolate similar lanthanide complexes with a
smaller ethylene spacer bisected by the idealized mirror
plane, we rationalized that the slightly larger propyl unit
would increase the requisite N‚‚‚N separation for larger
lanthanide-metal capture.6e Second, the avoidance of aromatic
groups within the ketoiminate backbone would preclude
potential intermolecular π-stacking interactions, which have
been implicated in the formation of anhydrous, dinuclear
Ln2L3 (L ) Schiff base)8a that are incapable of optimally
serving as single-site catalysts.
Ligand metalation is accomplished by reacting the protio
ketoiminate with Ln[N(TMS)2]3 in refluxing C7H16, the
silylamide route.12a,b The homoleptic amido starting materials
were prepared from dried Ln(OTf)3 using our previously
reported procedure.6e The use of “innocent” linear hydro-
carbons during metalation avoids metal ion solvation,
whereas the silylamide-mediated lanthanide insertion mutes
the potential formation of “ate” complexes or µ-bridging8b
halides from LnCl3 preparations.8b The as-isolated mono-
amidos are themselves active catalysts whose metal coordi-
nation spheres are not crowded by spurious ligands, the
HN(TMS)2 byproduct is easily removed from stoichiometric
reactions, and the monodentate ligand can be rapidly replaced
using acid-base chemistry without degradation of the Schiff-
base scaffold (vide infra). Although RajanBabu,9a Anwander,9b
and Evans6d have successfully employed both silylamide and
chloride routes for mononuclear and dinuclear yttrium salen
complexes that are catalytically active,9b the metal ion in
these examples is further stabilized through coordination to
a Lewis base (i.e., THF) from either adventitious solvent or
starting material solvates. This seemingly subtle solvate
increase in coordination number can drastically diminish the
catalytic activity.13a,b
give X-ray-quality crystals in excellent yield and analytical
purity. Although the ketoiminate framework does not degrade
in the presence of acidic phenols, ethanol, or free HN(TMS)2,
compounds 7 and 8 are completely inactiVe for the solution
polymerization of propylene oxide, ꢀ-caprolactone, butyro-
lactone, and rac-lactide. These results parallel those of
Hubert-Pfalzgraf, who also observed no polymerization
activity of polar monomers with trivalent rare earths coor-
dinated to bulky aryloxide monoanions.14 This is attributed
to substrate inaccessibility at the metal center and is further
supported by noting that the less sterically demanding
complex, 10, acts as an active polymerization catalyst for
these monomers. Analogously, we sought to assess requisite
alkoxide sterics for isolating discrete mononuclear complexes
by employing ethanol as the Brønsted acid (Scheme 1). This
extreme reduction in alcohol bulk afforded a neutral, trimeric,
oxo-alkoxide Er3+ cluster (11). Although the origin of the
bridging µ3-oxo dianion remains speculative, adventitious
H2O or oxo-atom abstraction are the most likely sources.15
The solid-state structure of 4 exhibits a mononuclear, five-
coordinate Sm3+ ion in a slightly distorted (τ ) 0.08)16
square-pyramidal environment. The average Sm-O and
Sm-N ketoiminato bond lengths17 are 2.218 (2,9,9,2) and
2.466 (2,6,6,2) Å, respectively, and the apical 2.307(2)-Å
Sm-N(3) amido bond distance 4.1° from the ketoiminato
N2O2 square base normal. The Sm3+ ion is displaced 0.83 Å
from the (0.05-Å S4-ruffled N(1), N(2), O(1), O(2) mean
plane. The canting of the apical amido and the Schiff-base
S4 ruffling impose chirality upon the metal center, and both
enantiomers are present within the crystal. The six-membered
diamino chelate ring, bearing the gem-dialkyl substituent at
C(6), adopts the chair conformation previously reported6c for
the isomorphous erbium complex. This is in contrast to the
unsubstituted six-membered diamine chelate rings in com-
plexes 5-10, all of which adopt the boat conformation
without crystallographic disorder at the C(6) methylene
carbon.18
To determine ketoiminato stability toward amido substitu-
tion and develop general routines to produce more nucleo-
philic monoanions (i.e., alkoxides), we accomplished direct
conversion of the mononuclear amidos to mononuclear
alkoxides by the stoichiometric reaction of 1, 5, and 6 with
sterically demanding 2,6-di(tert-butyl)-4-methyl phenol (BHT)
(14) Daniele, S.; Hubert-Pfalzgraf, L. G.; Vaissermann, J. Polyhedron 2003,
22, 127.
(15) Caulton, K. G.; Hubert-Pfalzgraf, L. G. Chem. ReV. 1990, 90, 969.
(16) For the τ calculation and its significance in five-coordinate square-
planar complexes, see: Addison, A. W.; Rao, T. N.; Reedijk, J.; van
Rijn, J.; Verschoor, G. C. J. Chem. Soc., Dalton Trans. 1984, 1349.
(17) Throughout this manuscript, a standard formalism of crystallography-
derived estimated standard deviations is used to provide a more
concise, yet comprehensive, synopsis of metrical parameters. Hence,
the first number in parentheses following an average value of a bond
length is the rms estimated standard deviation of an individual datum.
The second and third numbers are the average and maximum deviations
from the average value, respectively. The fourth number represents
the number of individual measurements included in the average values.
2a
in C5H12 (Scheme 1). The crude alkoxide complexes are
readily purified as innocent solvates by recrystallization to
(12) (a) Anwander, R. Top. Curr. Chem. 1996, 179, 33. (b) Anwander, R.;
Runte, O.; Eppinger, J.; Gerstberger, G.; Herdtweck, E.; Spiegler, M.
J. Chem. Soc., Dalton Trans. 1998, 847.
(13) (a) Hessen, B. Presented at the 225th National Meeting of the American
Chemical Society, New Orleans, LA, March 2003; paper INOR-919.
(b) Gagne, M. R.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1992,
114, 275.
6208 Inorganic Chemistry, Vol. 43, No. 20, 2004

Lanthanide Amido and Aryloxide Complexes
The mononuclear complexes with the unsubstituted propyl
bridge (5-10) have several common solid-state features. The
polyhedron in each, defined by the coordinated oxygen and
nitrogen atoms, is a distorted square pyramid, and the
ketoiminato donor atoms define an S4-ruffled square base.
The extent of S4 ruffling (∆L in Table 3) ranges from 0.009
Å in 7 to 0.19 Å in 5. Each Ln3+ ion is displaced (∆M in
Table 3) by 0.83-0.93 Å from this square base. The average
ketoiminato Er-O bond lengths range from 2.131 (2,1,1,2)
Å in 9 to 2.144 (3,18,18,2) Å in 10 and average 2.140
(2,21,9,12) Å for all six mononuclear propyl-bridged erbium-
containing molecules reported herein. The corresponding
ketoiminato Er-N bond lengths range from 2.365 (2,8,8,2)
Å in 9 to 2.371 (2,2,2,2) Å in 5 and average 2.369 (3,16,7,12)
Å for all six mononuclear erbium complexes. The corre-
sponding ketoiminato Sm-O and Sm-N bond lengths of
2.206 (2,0,0,2) and 2.464(3,3,3,2) Å in 8 are completely
consistent with the 0.07-Å larger ionic radius of Sm3+ relative
to Er3+. The axial Er-N(3) bonds range from 2.225 (2) Å
in 6 to 2.230 (2) Å in 5. The axial Er-O(3) bonds range
from 2.071 (3) Å in 10 to 2.081 (2) Å in 9. The unsubstituted
propyl bridge in each of these complexes is part of a six-
membered (-Ln-N1-C5-C6-C7-N2-) diamino chelate
ring that adopts an energetically unfavorable boat conforma-
tion. This chelate ring conformation produces a relatively
short 2.6-2.7-Å Ln‚‚‚H contact within each complex to a
diastereotopic hydrogen on C(6) and may represent a
stabilizing agostic interaction. It is particularly noteworthy
that this supposition is supported by a decrease of the 13C
JC(6)-H values when compared with the free ligand.19
The axial Ln-O(3) and Ln-N(3) bonds in all of the
mononuclear species make angles between 1.6-7.6° with
the normals to their square-base mean planes. This canting
( ax in Table 3) is the cumulative result of sterically induced
distortions of the prochiral complex from idealized Cs
symmetry and, in conjunction with the S4 ruffling, produces
a form of “conformational racemic chirality”.20
The solid-state structures of the six amido and aryloxide
complexes, 5-10, share many structural similarities, and their
salient metrical parameters are summarized in Table 2.
Although none of these molecules possess rigorous crystal-
lographic symmetry, each approximates prochiral Cs-m to a
varying degree. The idealized mirror plane would always
contain the lanthanide ion, the axially coordinated nitrogen
or oxygen atom (N(3) and O(3), respectively), and the C(6)
central carbon atom of the propyl bridge as well as any
substituents decorating this position. However, any structural
distortion that removes this symmetry element introduces
chirality at the metal center. Although this symmetry
reduction could occur in many ways, five appear to be
important in 5-10 because of geometric constraints inherent
to the ligand sets, these include: (1) rotation of the amido
or aryloxide ligands about the apical Ln-N or Ln-O bond,
respectively; (2) rotations of either the TMS group about
t
the N-Si bonds or the Bu groups about the R-ArC-tBu
bonds; (3) tilting of the axial Ln-N(3) or Ln-O(3) bonds
relative to the square-base normal. (4) dissimilar folding of
the two Schiff-base halves along the O‚‚‚N polyhedral edges
t
of this base; (5) rotation of the “frontal” Schiff base Bu or
Ph substituents about their KIC-tBu or KIC-Ph bonds.
Because each mononuclear, chiral complex utilizes a cen-
trosymmetric space group, both enantiomers are present in
equal amounts within any crystal. Furthermore, crystalline
6 contains two separate enantiomeric pairs. The superposition
of any enantiomeric pair would, therefore, show the extremes
of a given symmetry-breaking distortion upon the prochiral
species.
The structural constraints of the individual ligands in a
given complex determine the nature and extent of any
distortion from ideal Cs-m symmetry. Each Schiff-base ligand
contains two planar ketoiminato halves (both containing one
oxygen, one nitrogen, and six carbons) joined at a common
point, the central carbon, C(6), of the propyl bridge. These
planar halves can fold at the O‚‚‚N square polyhedral edge
and/or rotate about the vector through C(3) or C(9) and the
midpoint of the corresponding O‚‚‚N edge. S4 ruffling of
the square-base donor atoms is the structural manifestation
of either planar ketoiminato half rotating relative to the other.
The extent of Schiff base folding and/or rotation is restricted
by the stereochemical requirements of the central methylene,
C(6), whose six-membered diamine chelate ring always
adopts the thermodynamically unfavorable boat conforma-
tion. The observance of diastereotopic Ln‚‚‚H close contacts
(without disorder at C(6)) and a decrease in the 13C JC(6)-H
coupling constant compared to that in the free ligand suggests
an agostic interaction that conformationally “locks” this
stereochemical pivot point. This structural feature is not
observed in 2, 4, or the related propyl-bridged, mononuclear
RAl3+(Salpen) complexes of Atwood21 and Nomura.20
The mononuclear structures observed for 7-10 bearing
bulky apical aryloxides minimize nonbonded repulsions
t
between the Bu methyl groups, the apical oxygen, and the
adjacent meta ring hydrogens in the free ligand. This C2V
structure places two methyl groups from each tBu substituent
close to the oxygen atom, O(3), and the third methyl close
to the meta ring hydrogens. Similar alkyl group conforma-
tional orientation has been observed in homo- and hetero-
leptic lanthanide 2,6-di(tBu) phenoxide complexes.22,23 Steric
interactions between two of the methyl groups (C(28) and
C(33)) and the first and third methylene carbon atoms (C(5)
and C(7)) of the propyl bridge also force the aryloxide to
(18) Propyl-bridged Cu2+ Salen complexes have recently been found to
contain this folded geometry; however, the central methylene is
disordered over two positions with respect to the N2O2 plane, which
is not the case in our systems; see Nathan, L. C.; Koehne, J. E.;
Gilmore, J. M.; Hannibal, K. A.; Dewhirst, W. E.; Mai, T. D.
Polyhedron 2003, 22, 887.
(19) For the use of 1JC-H coupling constants as indicators for Ln3+‚‚‚H-C
interactions; see Gordon, J. C.; Giesbrecht, G. R.; Brady, J. T.; Clark,
D. L.; Keogh, D. W.; Scott, B. L.; Watkin, J. G. Organometallics
2002, 21, 127.
(20) Nomura, N.; Ishii, R.; Akakura, M.; Aoi, K. J. Am. Chem. Soc. 2002,
124, 5938. A particularly important consequence within this paper is
the possibility of “conformational racemic chirality” about the C3H6
bridge of an achiral aluminum Salen, thereby facilitating the site-
controlled mechanism of isotactic polylactide formation.
(21) Atwood, D. A.; Harvey, M. J. Chem. ReV. 2001, 101, 37. (b) Atwood,
D. A.; Hill, M. S.; Jegier, J. A.; Rutherford, D. Organometallics 1997,
16, 2659.
Inorganic Chemistry, Vol. 43, No. 20, 2004 6209

Schuetz et al.
those found in other five-coordinate erbium(III)22 and sa-
marium(III)17,24 heteroleptic complexes, and their aforemen-
tioned angles are slightly more acute. Although the expla-
nations17,25 behind these observations remain debatable,
agostic interactions between the Ln3+ ions and the 2,5-tBu
aryloxide substituent hydrogens are almost certainly not
uniquely responsible for this nearly linear Ln-O(3)-C(20)
bond (vide infra), and neither complex exhibits noteworthy
spectral evidence of Ln‚‚‚H interactions.
Unlike the (KI)Ln(BHT) complexes, an invariant structure
is not expected or observed for the (KI)Ln(N(TMS)2)
complexes. A conformation having one methyl from each
TMS group close to the apical nitrogen atom would,
however, be sterically favorable for the free ligand, and this
arrangement is usually observed in its lanthanide complexes.
Although the totally eclipsed conformation would possess
C2V symmetry, it would simultaneously introduce unfavorable
nonbonded methyl‚‚‚methyl contacts between the two TMS
groups. These repulsions are minimized by slight rotation
about the Si-N bonds while still leaving one methyl from
each TMS (C(25) and C(29), Figure 4), considerably closer
to N(3) than the other two methyls. The coordination of such
an amido ligand places these two methyl groups over the
Ln-O-C-C-C-N- chelate rings near the O‚‚‚N poly-
hedral-edge midpoints. Both of these amino methyls protrude
further toward the N2O2 square and/or Schiff base than the
four BHT methyls, which are directed toward the ketoiminato
halves (∆Me in Table 3). Therefore, the amino ligands should
produce a more pronounced folding of the Schiff base along
the O‚‚‚N polyhedral edges and/or rotations about the vectors
between C(3) or C(9) and the corresponding O‚‚‚N mid-
points. Furthermore, the TMS groups are not locked by the
steric volume of the axial ligand near the metal or by the
nonbonded contacts with the bridging C(5) and C(7) meth-
ylenes. Thus, the amido ligand has more freedom to sample
alternate conformations. For example, the entire amido ligand
can oscillate about the Ln-N(3) bond while minimizing any
methyl-ketoiminato repulsions through a combination of
dissymmetric ketoiminato folding or TMS group rotation
about the N-Si bonds. This flexibility also allows for greater
canting of the Ln-N(3) bond relative to the square-base
normal as well as its tipping toward the propyl bridge to
give a more accessible metal center for (KI)Ln[N(TMS)2]
complexes relative to the corresponding (KI)Ln[BHT] spe-
cies.
Figure 2. Molecular structure of mononuclear (KI)Sm3+ aryloxide (8)
that is isomorphous with 7.
adopt the general orientations shown in Figure 2 and restrict
rotation about the Ln-O(3) bond. Thus, four methyl groups
(C(27), C(28), C(32), and C(33)) impinge downward upon
the ketoiminato CdO and C-N bonds and afford reasonably
symmetrical Schiff-base folding along the O‚‚‚N polyhedral
edges. This bilateral steric effect is manifested in both
decreased S4 ruffling in the KI donor atoms and smaller τ
values when compared to the parent aminos (Table 3).
Unfortunately, these four methyl groups render the metal ion
inaccessible to the incoming substrate and contribute to the
complete solution-based polymerization inactivity toward
strained (i.e., epoxides) and unstrained (i.e., lactones and
lactides) polar monomers. Although much of the (BHT)Ln-
(KI) is sterically locked into a specific conformation, the
Ln-O(3) bond cants off the square-base normal in the plane
of its phenyl ring; the frontal ketoiminato substituents
accommodate this motion by adopting different relative
orientations. This is evident from the front (Figure 3A) and
side (Figure 3B) views of the superimposed enantiomers of
7. The metal and Schiff-base nonhydrogen atoms (excluding
methyl carbons) for the racemic pair were superimposed by
least-squares fitting to give a weighted rms deviation of 0.03
Å and an O(3)-Er-(O3′) angle of 14.1°.
The apical metal aryloxide bonds in 7-10 are significantly
(∼0.06 Å) shorter than the ketoiminato Ln3+-O bonds, as
expected on the basis of charge density. The three indepen-
dent Er-O(3) bond lengths range from 2.071(3) to 2.081(2)
Å, and the Sm-O(3) bond length in 8 is 2.143(2) Å; these
apical Ln-O(3) bonds are all canted (Table 3) from the
square-base normal. The Sm-O(3)-C(20) bond angle is
173.1(2)°, and the Er-O(3)-C(20) bond angles range from
172.6(3) to 175.6(1)°. Considering the intense recent interest
in rare-earth alkoxide (and chalcogenide) bonding, the Ln3+-
O(3) bond lengths in 7-10 reside in the “shorter” range of
This rotational variability about the Ln-N(3) bond is
clearly evident from the plan views (Figure 4) for the two
crystallographically independent molecules present in 6. Even
though these two molecules are approximately related in the
crystal by a noncrystallographic inversion center at (1/4, 0,
1/4) in the unit cell, they clearly differ in the extent of rotation
about the Ln-N(3) bond. In comparison, the appropriately
inverted metal and 15-nonhydrogen atom (vide supra) Schiff-
base frameworks for these two crystallographically indepen-
dent molecules superimpose with a weighted rms deviation
(22) Hitchcock, P. B.; Lappert, M. F.; Singh, A. J. Chem. Soc., Chem.
Commun. 1983, 1499.
(23) Clark, D. L.; Gordon, J. C.; Huffman, J. C.; Vincent-Hollis, R. L.;
Watkin, J. G.; Zwick, B. D. Inorg. Chem. 1994, 33, 5903.
(24) Katagiri, K.; Kameoka, M.; Nishiura, M.; Imamoto, T. Chem. Lett.
2002, 426.
(25) Russo, M. R.; Kaltsoyannis, N.; Sella, A. Chem. Commun. 2002, 2458.
6210 Inorganic Chemistry, Vol. 43, No. 20, 2004

Lanthanide Amido and Aryloxide Complexes
Figure 3. Superimposed enantiomers of complex 7 showing canting of the apical Er-O bond within the phenyl-ring plane.
Figure 4. Enantiomers of complex 6 exhibiting Er-N bond twisting.
of 0.30 Å and a N(3)-Er-N(3′) angle of 5.1°. The metal
and 15-nonhydrogen atom SB framework for the enantiomer
of molecule 1 superimposes with a 0.35-Å rms deviation
onto molecule 1 and gives a Nap-Er-Nap′ angle of 0°. A
similar superposition for the enantiomer of molecule 2 on
molecule 1 gives a 0.10-Å rms deviation and a Nap-Er-
Nap′ angle of 5.0°, which indicates that the steric effects of
the apical N(TMS)2 on the Schiff-base distortion are a
function of amido ligand tilting and rotation about the amido
Ln-N bonds and Si-N bonds.
10. Accordingly, 7 exhibits the smallest square-base ruffling
((0.01 Å) when compared to those of 5 and 6 and 8-10,
whereas 10 has moderate S4 ruffling of (0.14 Å. This can
t
be attributed to four locked Bu methyl groups (two pairs
on each side of the molecule) buttressed against each other
at the front of the molecule in 7 but not in 10 because phenyl
t
groups have replaced the Bu groups. The frontal phenyl
substituents’ orientation is nearly coplanar with the planar
ketoiminate halves in our complexes (dissymmetric left to
right phenyl-ketoiminato dihedral angles in 10 are 9.1 and
24.1°, respectively) as well as in a recently reported Pd2+
dinuclear complex bearing a pentadentate Schiff base derived
from the identical diketone.26 This structural arrangement
produces notably short H...H contacts between an ortho
hydrogen (on C(16) or C(23)) and the adjacent KI methine
hydrogen (on C(3) or C(9)). This is presumably not a
sterically imposed conformation but rather is attributable to
favorable π-π interactions between the phenyl ring and the
delocalized KI π system. The nearly coplanar arrangement
of the phenyl substituents greatly reduces steric crowding at
the front of the molecule and is most likely responsible for
the observation that complex 10 is an active catalyst for the
ring opening of lactides, lactones, and strained epoxides,
whereas 7-9 fail to react with these monomers.
The structures of 7 and 10 provide a basis for comparing
the structural effects of different frontal KI substituents. Both
compounds have an apically coordinated, anionic BHT ligand
and differ only in the frontal tBu (7) and Ph (10) substituents.
The frontal tBu substituents have the same rotational restric-
t
tions as the BHT Bu groups, and they too prefer to have
two methyls “straddle” the adjacent carbonyl oxygen with
limited rotational freedom about the KIC-C(CH3)3 bond. The
metal and 15-nonhydrogen atom SB framework for the
enantiomers of 7 superimposed with a 0.03-Å rms deviation
and a 14.1° Oap-Er-Oap′ angle, whereas the enantiomers
of 10 superimposed (Figure 5) with a 0.40-Å rms deviation
and a 13.6° Oap-Er-Oap′ angle. It is apparent from Figures
3 and 5 that the Er(KI) fragment differences across the
molecular pseudo-mirror plane are almost nonexistent in 7
but become much larger toward the front of the molecule in
(26) Mikuriya, M.; Minowa, K.; Lim, J.-W. Bull. Chem. Soc. Jpn. 2001,
74, 331.
Inorganic Chemistry, Vol. 43, No. 20, 2004 6211

Schuetz et al.
Figure 5. Enantiomers of 10 superposed with a 0.40 Å rms deviation and a 13.6° Oap-Er-Oap′ angle. As aryloxide canting occurs, the frontal phenyl
substituents rotate to become as coplanar as possible with the planar ketoiminate halves.
Figure 6. Neutral trimeric Er3+ complex 11 (A) with a central core composed of three seven-coordinate Er3+ ions bridged by µ3-oxo, µ3-alkoxide, and
µ2-O (ketoiminato) linkages (B). Each Er3+ ion displays monocapped trigonal prismatic coordination geometry (C), and the individual subunits share capping
triangles via O(1)‚‚‚O(1S) edges in the assembled polyhedra representation (D).
One additional structural feature of alkoxides 7-10 needs
to be discussed. Tilting of the alkoxide ligands in a plane
that is perpendicular to the molecular pseudomirror produces
3.01 Å persist in 7. With the methyl hydrogens being
positioned as freely rotating rigid rotors with C-H bond
lengths of 0.96 Å, these contacts would be shorter if more
realistic C-H bond lengths (i.e., longer) were employed and
the methyl groups were rotated for maximal Er‚‚‚H interac-
tion. Even so, the observed 2.84-Å Er‚‚‚H distance in 7 is
only 0.14 Å longer than the standard Er‚‚‚H distance
involving methylene carbon C(6) in 5-10.
t
short (e3.05 Å) Ln...H contacts involving the ortho Bu
methyl groups (C(27),C(28),C(32),C(33)). Although apical
ligand tilting is obvious (Figures 4 and 6), there is no direct
spectroscopic evidence for formal M‚‚‚H interactions with
the methyl groups. Nonetheless, Er‚‚‚H contacts of 2.84 and
6212 Inorganic Chemistry, Vol. 43, No. 20, 2004

Lanthanide Amido and Aryloxide Complexes
Although slightly decreasing ketoiminato frontal steric bulk
(tBu to Ph) still affords mononuclear aryloxide complexes,
the stoichiometric alcoholysis of 6 with the significantly
smaller ethanol reproducibly yields a neutral trimeric cluster,
11 (Figure 6A), whose central core consists of three seven-
coordinate Er3+ ions bridged by µ3-oxo, µ3-alkoxide, and
µ2-O (ketoiminato) linkages (Figure 6B). The metal ions
remarkably serve as the vertices of a nearly perfect equilateral
triangle with an edge length of 3.324 (1,7,5,3) Å. Similar,
less-ordered dicationic and tricationic species have been
higher temperatures, resulting from weaker association of
the metal ion and growing polymer chain. For ꢀ-caprolactone
under similar conditions, Mn (103) falls between 33.9 and
38.7 with monomodal polydispersities displaying a narrower
range of 2.0 to 3.2. Although these preliminary results
establish polymer initiation, a more detailed understanding
of the polymerization process, capture of the initiating
species, and more controlled polymerizations will be forth-
coming.
Summary
27
reported from aqueous reactions, including Gd3(µ3-OH)2
and La3(µ3-OH)(µ2-H2O)28 cores, respectively, as their neutral
nitrato salts. The average metrical parameters for the three
types of ketoiminato donor atoms in 11, Er-N, Er-O, and
Er-O(µ2), are 2.428 (3,20,8,6), 2.211 (3,9,6,3), and 2.322
(2,30,19,6) Å, respectively. The average distances of the
common Er-O2-(µ3) and Er-OEt1-(µ3) are 2.187 (2,6,4,3)
and 2.417 (2,20,14,3) Å, respectively. The ethoxide CH3
terminus is disordered over two positions (modeled at 54
and 46% occupancies) within clefts created by the Schiff-
base phenyl groups, all of which reside on a single side of
the molecule. The coordination geometry about each Er3+
ion is best described as a monocapped trigonal prism29
(Figure 6C), whose triangular faces are only 1.3° from being
parallel. Three of these monocapped trigonal prismatic
subunits subsequently assemble with pseudo C3 symmetry
into the trinuclear cluster. Each of these subunits shares a
pair of capping triangles having an O(1)‚‚‚O(1S) edge with
adjacent polyhedra (Figure 6B and D). The square face
defined by O(1), O(11), O(12), and O(32) exhibits 0.09-Å
S4 ruffling (note: this is not related to the aforementioned
discussion of the square planar complexes) with the capping
O(1S) atom displaced 1.83 Å from the square mean plane.
Synthetically, 11 immediately decomposes upon exposure
to the ambient atmosphere (even bathed in Paratone-N),
necessitating that the crystals be exclusively handled and
mounted at low temperatures in a glovebag.
Initial solution-based homopolymerizations30 with com-
pounds 1, 4-6, and 10 and purified ꢀ-caprolactone and rac-
lactide monomers afford biodegradable polyesters. For the
amidos, 1 and 4-6, and the alkoxide, 10, at [monomer]/
[catalyst] loads of 150:1 in toluene at 26 and 70 °C for 30
min, high-molecular weight polymers (>92% conversion by
1H NMR and TGA analyses) exhibiting broad polydisper-
sities were recovered. At 26 °C, for both atactic and isotactic
lactide polymers, Mn (103) varies from 5.3 to 12.6 (tacticity
depends on catalyst structure) and exhibits polydispersities
from 1.4 to 2.6. At 70 °C, Mn (103) ranges from 3.6 to 15.6,
again depending on catalyst structure; however, the poly-
dispersities broaden from 4.3 to 6.2. These preliminary
observations are attributed to increased chain transfer at
This contribution presents the syntheses of mononuclear,
anhydrous, solvent-free, five-coordinate lanthanide amido
complexes 4-6 bearing tetradentate Schiff bases. The Schiff-
base ligands are devoid of backbone aromatic groups and
contain a C3H6 spacer between the ketoiminato halves,
providing sufficient N‚‚‚N separation for larger lanthanide
metal capture. Three organic scaffolds, with different front
(tBu and Ph) and rear (gem-dimethyl and methylene) steric
requirements, all successfully support low coordination
numbers (<6) spanning seven contiguous f-element Ln3+.
Direct conversion, in high yields and analytical purity, of 1,
5, and 6 to bulky aryloxides (7-10) is accomplished by the
stoichiometric reaction with 2,6-bis(tert-butyl)-4-methylphe-
nol, without degradation of the N2O2 ligand framework but
with solvent polymorphism. Attempts to prepare analogous
species with a less sterically demanding alcohol, EtOH,
reproducibly afforded the neutral trinuclear complex, 11.
A summary of salient mononuclear amino and alkoxide
metrical parameters is given in Tables 2 and 3. The five-
coordinate complexes, 5-10, share similarities, including:
(1) boat conformation for the six-membered diamine chelate
ring that establishes a close contact between the Ln3+ ion
and a diastereotopic hydrogen on the central C(6) methylene;
(2) square pyramidal metal coordination with the Schiff-base
donor atoms defining the square base; (3) varying degrees
of S4 ruffling (∆L) for the square-base donor atoms; and (4)
canting ( ax) of the apical Ln-N(3) or Ln-O(3) bond
relative to the normal to this square base. From Table 2, it
is clear that the greatest degree of N2O2 ruffling and the
largest τ values occur for the amido complexes, 5 and 6;
these complexes also have significant asymmetrical folding
along the O-N KI edges. This creates innate chirality at
the metal center, affording both enantiomers (or racemic
pairs) within the crystals and is a result of both KI flexibility
and dissymmetric KI‚‚‚CH3[Si(CH3)2] steric repulsions for
the two TMS methyl groups (on either side of the idealized
σv) protruding toward the square base. Conversely, the
t
symmetrical orientation of four aryloxide Bu, CH3 groups
downward upon KI C-N and CdO linkages affords more
symmetrical ligand folding. Thus, nearly ideal square-pyr-
amidal geometry is observed for the sterically hindered
aryloxide complex 7 and its solvent polymorph 9, despite
different solvent-induced packing arrangements.
Complexes 1, 4-6, and 10 are active initiators for solution-
based homopolymerizations of rac-lactide and ꢀ-caprolactone
at temperatures of 26 and 70 °C. Initial experiments provided
polymer samples with moderate molecular weights and
(27) Costes, J.-P.; Dahan, F.; Nicode`me, F. Inorg. Chem. 2001, 40, 5285.
(28) Aspinall, H. C.; Black, J.; Dodd, I.; Harding, M. M.; Winkley, S. J.
J. Chem. Soc., Dalton Trans. 1993, 709.
(29) Kepert, D. L. Prog. Inorg. Chem. 1979, 25, 41.
(30) (a) Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 2002, 124, 1316.
(b) Aubrecht, K. B.; Hillmyer, M. A.; Tolman, W. B. Macromolecules
2002, 35, 644. (c) O’Keefe, B. J.; Monnier, S. M.; Hillmyer, M. A.;
Tolman, W. B. J. Am. Chem. Soc. 2001, 123, 339. (d) Ovitt, T. M.;
Coates, G. W. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4686.
Inorganic Chemistry, Vol. 43, No. 20, 2004 6213

Schuetz et al.
polydispersities that broaden at higher temperatures. Under
identical conditions, 7-9 are completely inactive for the
solution polymerization of strained epoxides, lactides, and
lactones, presumably because of the monomer’s inacces-
sibility to the metal center. As expected, active polymer
initiation accompanies decreased frontal steric bulk (9 vs 10).
of NebraskasLincoln is also acknowledged for assistance
in obtaining the paramagnetic NMR data presented in this
manuscript.
Supporting Information Available: Crystallographic treatment
of hydrogen atoms and disordered solvent and thermal ellipsoid
plots (where ball-and-stick models are used) and crystallographic
files for 4-11 in CIF format. This material is available free of
charge via the Internet at http://pubs.acs.org. These may also be
obtained from the Cambridge Crystallographic Data Center, 12
Union Road, Cambridge, CB2 1EZ, U.K. (fax: +44-1223-
336033; e-mail: deposit@ccdc.cam.ac.uk) under assigned CCDC
numbers (compound) of 210063 (4), 210064 (5) 210067 (6), 210065
(7), 210066 (8), 210070 (9), 210069 (10), and 210068 (11).
Acknowledgment. This work was supported by the
University of Nebraska Layman Foundation and the Ne-
braska Research Initiative, and acknowledgment is also made
to the donors of the American Chemical Society Petroleum
Research Fund for supporting this research. J.A.B. also
thanks Professor Eric Fossum of Wright State University for
numerous polymerization discussions and for obtaining
preliminary GPC data. Dr. Dipanjan Nag of the University
IC040006F
6214 Inorganic Chemistry, Vol. 43, No. 20, 2004