An yttrium-based system to evaluate Lewis base coordination to an lectropositive metal in a metallocene environment

Article abstract of DOI:10.1021/om010908d

The unsolvated bimetallic yttrium complex (C₅Me₅)₂Y(μ-Cl)Y(C₅Me₅ )₂Cl (1) provides a convenient platform upon which to compare the coordination chemistry of oxygen-donor ligands and monomers with Lewis acidic metal ions. Reaction of 1 with 2 equiv of oxygen-containing substrates formed the monomeric complexes (C₅Me₅)₂YCl(L) (L = THF (2), benzophenone (3), methyl methacrylate (4), ε-caprolactone (5), hexamethylphosphoramide (6), ε-caprolactam (7), 1-methyl-2-pyrrolidinone (8), N,N′-dimethylpropyleneurea (9)). Each of these readily crystallize, which allows comparison of the Y-O interaction in the solid state. The X-ray data show that the numerical values of the measured Y-O lengths decrease in the order THF > benzophenone > methyl methacrylate > ε-caprolactone > hexamethyl-phosphoramide > ε-caprolactam > 1-methyl-2-pyrrolidinone > N,N′-dimethylpropyleneurea, although these bond lengths span only a short range and some are the same within experimental error. In the case of e-caprolactam, a bis adduct, (C₅Me₅)₂YCl(ε-caprolactam)₂ (10), can be isolated from the reaction of 1 and excess ε-caprolactam. ¹H NMR spectroscopy indicates that (C₅Me₅)₂YCl(L)L′ ligand displacement reactions in solution with 2-9 follow a trend consistent with the bond lengths except for (C₅Me₅)₂YCl(ε-caprolactam) (7), in which L participates in hydrogen bonding to the chloride.

Full text of DOI:10.1021/om010908d

Organometallics 2002, 21, 1825-1831
1825
An Yttr iu m -Ba sed System to Eva lu a te Lew is Ba se
Coor d in a tion to a n Electr op ositive Meta l in a
Meta llocen e En vir on m en t
William J . Evans,* Cy H. Fujimoto, Matthew A. J ohnston, and J oseph W. Ziller
Department of Chemistry, University of California, Irvine, Irvine, California 92697-2025
Received October 17, 2001
The unsolvated bimetallic yttrium complex (C5Me5)2Y(µ-Cl)Y(C5Me5)2Cl (1) provides a
convenient platform upon which to compare the coordination chemistry of oxygen-donor
ligands and monomers with Lewis acidic metal ions. Reaction of 1 with 2 equiv of oxygen-
containing substrates formed the monomeric complexes (C5Me5)2YCl(L) (L ) THF (2),
benzophenone (3), methyl methacrylate (4), ꢀ-caprolactone (5), hexamethylphosphoramide
(6), ꢀ-caprolactam (7), 1-methyl-2-pyrrolidinone (8), N,N′-dimethylpropyleneurea (9)). Each
of these readily crystallize, which allows comparison of the Y-O interaction in the solid
state. The X-ray data show that the numerical values of the measured Y-O lengths decrease
in the order THF > benzophenone > methyl methacrylate > ꢀ-caprolactone > hexamethyl-
phosphoramide > ꢀ-caprolactam > 1-methyl-2-pyrrolidinone > N,N′-dimethylpropyleneurea,
although these bond lengths span only a short range and some are the same within
experimental error. In the case of ꢀ-caprolactam, a bis adduct, (C5Me5)2YCl(ꢀ-caprolactam)2
1
(10), can be isolated from the reaction of 1 and excess ꢀ-caprolactam. H NMR spectroscopy
indicates that (C5Me5)2YCl(L)/L′ ligand displacement reactions in solution with 2-9 follow
a trend consistent with the bond lengths except for (C5Me5)2YCl(ꢀ-caprolactam) (7), in which
L participates in hydrogen bonding to the chloride.
In tr od u ction
binding preference of monomers that contain several
potential oxygen donors in a polyfunctional substrate.
We report here our attempts to develop a system that
can be conveniently used to evaluate the relative donor
capacity of oxygen-containing substrates to metals such
as yttrium and the lanthanides in typical organometallic
environments. Our data suggest that (C5Me5)2Y(µ-
Cl)Y(C5Me5)2Cl (1) is appropriate for this purpose, since
it is readily synthesized9 and diamagnetic and it reacts
in high yield with oxygen donor complexes, L, to make
(C5Me5)2YCl(L) adducts. The adducts crystallize to give
one of the more common types of organometallic coor-
dination environments: i.e., one involving two (C5Me5)-
ligands, an anionic ligand (Cl)-, and a neutral donor
ligand. This system allows solid-state evaluation of the
Lewis acid-base interaction and competitive binding
studies via 1H NMR spectroscopy. The coordination
chemistry of 1 with eight oxygen-containing substrates
is reported here.
The large, electropositive, oxophilic yttrium and lan-
thanide metal ions frequently form complexes with
oxygen-donor ligands. Commonly this involves oxygen-
containing solvents such as THF,1-4 diethyl ether, and
dimethoxyethane (DME), but it has also been shown
that polymerizable monomers such as ꢀ-caprolactone5,6
and ꢀ-caprolactam7,8 can form strongly bound adducts
with the lanthanides. Although there are many ex-
amples of crystallographically characterized yttrium
and lanthanide metal complexes of oxygen-donor ligands,
less information is available on solid-state and solution
interactions of oxygen-containing polymerizable mono-
mers. Data on the hierarchy of coordination strengths
of oxygen-donor ligands would aid in choosing the proper
solvents for reactions involving oxygen-containing mono-
mers and in copolymerization systems. It would also be
useful to have details of the site of coordination and
(1) Evans, W. J .; Grate, J . W.; Levan, K. R.; Bloom, I.; Peterson, T.
T.; Doedens, R. J .; Zhang, H.; Atwood, J . L. Inorg. Chem. 1986, 25,
3614-3619.
Exp er im en ta l Section
(2) Mingqing, C.; Guang, W.; Shanming, Z.; Zuen, H.; Wenjie, Q.;
Wenling, W. Wuji Huaxue Xuebao 1986, 2, 102-104.
(3) Sobota, P.; Utko, J .; Szafert, S. Inorg. Chem. 1994, 33, 5203-
5206.
All of the compounds described below were handled under
nitrogen with rigorous exclusion of air and water using
standard Schlenk, vacuum line, and glovebox techniques.
Solvents were prepared, and physical measurements were
obtained as previously described.10 ꢀ-Caprolactone, methyl
methacrylate (MMA), N,N′-dimethylpropyleneurea (DMPU),
1-methyl-2-pyrrolidinone (MPU), and hexamethylphosphora-
(4) Zuo-Wei, X.; Chang-Tao, Q.; J ie, S.; Xiang-Lin, J . Chin. J . Struct.
Chem. 1993, 12, 107-110.
(5) Evans, W. J .; Shreeve, J . L.; Doedens, R. J . Inorg. Chem. 1993,
32, 245-246.
(6) Evans, W. J .; Shreeve, J . L.; Ziller, J . W. Inorg. Chem. 1995, 34,
576-585.
(7) Isolani, P. C.; Alvarez, H. A.; Matos, J . R.; Vicentini, G.;
Castellano, E.; Zukerman-Schpector, J . J . Coord. Chem. 1998, 43, 349-
360.
(9) Evans, W. J .; Peterson, T. T.; Rausch, M. D.; Hunter, W. E.;
Zhang, H.; Atwood, J . L. Organometallics 1985, 4, 554-559.
(10) Evans, W. J .; Chamberlain, L. R.; Ulibarri, T. A.; Ziller, J . W.
J . Am. Chem. Soc. 1988, 110, 6432-6432.
(8) Evans, W. J .; Fujimoto, C. H., Greci, M. A.; Ziller, J . W. Eur. J .
Inorg. Chem. 2001, 745-749.
10.1021/om010908d CCC: $22.00 © 2002 American Chemical Society
Publication on Web 03/30/2002

1826 Organometallics, Vol. 21, No. 9, 2002
Evans et al.
mide (HMPA) were purchased from Aldrich, dried over acti-
vated 3A molecular sieves, and degassed with three freeze-
pump-thaw cycles before use. ꢀ-Caprolactam was purchased
from Aldrich and sublimed before use. Benzophenone was
purchased from Fisher and used as received. (C5Me5)2YCl(µ-
Cl)Y(C5Me5)2 (1) was prepared as previously described.9 (C5-
Me5)2YCl(THF)1 was synthesized by a modified procedure as
described below. 1H and 13C NMR spectra were obtained on
an Omega 500 MHz or a GN 500 MHz spectrometer at 25 °C.
X-ray crystallographic data were obtained on a Bruker CCD
platform diffractometer. IR spectra were taken on thin films
obtained from benzene solutions (except where noted) using
an ASI ReactIR 1000 spectrometer. Complexometeric analyses
were obtained as previously described.11
(C6D6): 181.9 (CO), 117.0 (C5Me5), 42.5 (CH2), 36.4 (CH2), 30.0
(CH2), 28.3 (CH2), 22.6 (CH2), 11.4 (C5Me5). IR: 3205 m, 2927
s, 2858 s, 1630 vs, 1498 m, 1437 s, 1359 m, 1259 m, 1205 m,
1089 s, 1020 s, 803 s cm-1
.
(C5Me5)2YCl(1-m eth yl-2-p yr r olid in on e) (8). A white solid
(47.6 mg, 96%) was obtained from 1-methyl-2-pyrrolidinone
(10 mg, 0.10 mmol) and 1 (40 mg, 0.05 mmol). Anal. Calcd for
C25H39ClNOY: Y, 18.0. Found: Y, 17.5. 1H NMR (C6D6): δ 2.25
(s, 3H), 2.11 (s, 30H), 1.64 (m, 2H), 1.42 (m, 2H), 0.98 (m, 2H).
13C NMR (C6D6): δ 176.6 (CO), 116.6 (C5Me5), 49.9 (CH2), 31.5
(CH2), 30.2 (CH2), 17.4 (CH3), 1164 (C5Me5). IR: 2904 s, 1645
vs, 1514 m, 1444 m, 1313 m, 1259 m, 1112 m, 1027 m, 803 w,
765 w cm-1
.
(C5Me5)2YCl(DMP U) (9). A white solid (54 mg, 90%) was
isolated from DMPU (15 mg, 0.12 mmol) and 1 (45 mg, 0.056
mmol). Anal. Calcd for C26H42ClN2OY: Y, 17.0. Found: Y, 16.8.
1H NMR (C6D6): δ 2.82 (s, 3H), 2.45 (s, 3H), 2.22 (m, 2H), 2.18
(m, 2H), 2.15 (s, 30H), 0.91 (m, 2H). 13C NMR (C6D6): δ 157.3
(CO), 116.7 (C5Me5), 47.6 (CH2), 47.0 (CH2), 38.3 (CH3), 36.6
(CH3), 20.6 (CH2), 11.8 (C5Me5). IR: 2920 s, 2858 s, 1576 vs,
Typ ica l P r oced u r e for th e Syn th esis of (C5Me5)2YCl-
(L). P r ep a r a tion of (C5Me5)2YCl(THF ) (2). In a glovebox,
THF (14 mg, 0.20 mmol) in 5 mL of benzene was added to a
slurry of 1 (30 mg, 0.038 mmol) in about 1 mL of benzene.
The slurry instantly became a colorless solution. After 1 h of
stirring, the solvent was removed by rotary evaporation and
the white solid, 2, was obtained (35 mg, 80%) and identified
1413 w, 1321 w, 1259 s, 1020 vs, 905 w, 865 w, 766 vs cm-1
.
1
by H NMR spectroscopy.1
(C5Me5)2YCl(E-ca p r ola cta m )2 (10). The reaction of excess
ꢀ-caprolactam (60 mg, 0.53 mmol) with 1 (70 mg, 0.089 mmol)
gave a white solid which crystallized with one caprolactam in
the lattice, (C5Me5)2YCl(ꢀ-caprolactam)2‚(caprolactam) (125 mg,
96%). Anal. Calcd for C38H63ClN3O3Y: Y, 12.1. Found: Y, 12.3.
1H NMR (C6D6): δ 8.17 (b, 3H, ∆ν1/2 ) 49 Hz), 2.50 (m, 6H),
2.14 (m, 6H), 2.11 (s, 30H), 1.30 (m, 6H), 1.15 (m, 6H), 1.03
(m, 6H). 13C NMR (C6D6): δ 179.5 (CO), 116.5 (C5Me5), 42.4
(CH2), 36.8 (CH2), 30.4 (CH2), 29.5 (CH2), 23.2 (CH2), 11.7
(C5Me5). IR: 3290 w, 3205 m, 3051 w, 2927 s, 2858 s, 1637 vs,
1483 m, 1437 m, 1359 m, 1197 m, 1120 m, 1020 m, 981 w, 811
(C5Me5)2YCl(L) (L ) Ben zop h en on e (3), MMA (4), E-Ca -
p r ola cton e (5), HMP A (6), E-Ca p r ola cta m (7), 1-Meth yl-
2-p yr r olid in on e (8), DMP U (9)). 3-9 were prepared as
described for 2. Reagents, yields, and spectroscopic data for
individual samples are given below.
(C5Me5)2YCl(ben zop h en on e) (3). A dark red solid (40 mg,
91%) was isolated from benzophenone (14 mg, 0.08 mmol) and
1 (30 mg, 0.038 mmol). Anal. Calcd for C33H40ClOY: Y, 15.4.
Found: Y, 15.1. 1H NMR (C6D6): δ 7.76 (b, 4H, ∆ν1/2 ) 100
Hz), 7.10 (m, 2H), 7.04 (m, 4H), 2.04 (s, 30H). 13C NMR
(C6D6): δ 206.6 (CO), 135.1 (o-C), 132.3 (m-C), 129.3 (p-C),
117.8 (C5Me5), 11.8 (C5Me5). IR: 3059 w, 2912 s, 2858 s, 1614
s, 1568 s, 1444 s, 1328 s, 1290 s, 1159 m, 1081 s, 1027 s, 942
m, 803 m, 703 s cm-1. UV-vis (hexanes; λmax, nm (ꢀ, M-1cm-1)):
457 (1460).
m cm-1
.
X-r a y Da ta Collection , Str u ctu r e Deter m in a tion , a n d
Refin em en t. (C5Me5)2YCl(ben zop h en on e) (3). Complexes
3-10 were handled as described in detail here for 3, except
for the specific comments which follow. A red crystal of
approximate dimensions 0.15 × 0.28 × 0.49 mm was mounted
on a glass fiber and transferred to a Bruker CCD platform
diffractometer. The SMART12 program package was used to
determine the unit-cell parameters and for data collection (20
s/frame scan time for a sphere of diffraction data). The raw
frame data were processed using SAINT13 and SADABS14 to
yield the reflection data file. Subsequent calculations were
carried out using the SHELXTL15 program. The diffraction
symmetry was 2/m, and the systematic absences were consis-
tent with the centrosymmetric monoclinic space group P21/c,
which was later determined to be correct. Details are given in
Table 1.
The structure was solved by direct methods and refined on
F2 by full-matrix least-squares techniques. Analytical scatter-
ing factors16 for neutral atoms were used throughout the
analysis. Hydrogen atoms were included using a riding model.
(C5Me5)2YCl(E-ca p r ola cton e) (5). The pentamethylcyclo-
pentadienyl and caprolactone ligands exhibited a high degree
of thermal motion. The ring defined by C(11)-C(15) was
refined with isotropic thermal parameters using multiple
components with partial site-occupancy factors (0.60/0.40
major/minor components). The disordered atoms of the capro-
lactone ligand were also handled in this manner. It was
necessary to include bond distance constraints for this ligand.
(C5Me5)2YCl(MMA) (4). A yellow solid was isolated (50 mg,
94%) from MMA (10 mg, 0.10 mmol) and 1 (43 mg, 0.054
mmol). Anal. Calcd for C25H38ClO2Y: Y, 18.0. Found: Y, 18.2.
1H NMR (C6D6): δ 6.19 (m, 1H), 5.20 (m, 2H), 3.56 (s, 3H),
2.02 (s, 30H), 1.56 (s, 3H). 13C NMR (C6D6): δ 175.5 (CO),-
134.7 (CCdC), 131.8 (CH2), 117.6 (C5Me5), 56.8 (OCH3), 17.8
(CH3), 11.7 (C5Me5). IR: 2920 vs, 2858 vs, 1661 s, 1630 s, 1560
m, 1444 s, 1375 m, 1328 s, 1259 s, 1089 s, 1020 vs, 803 vs
cm-1
.
(C5Me5)2YCl(E-Ca p r ola cton e) (5). A white solid (49 mg,
94%) was obtained from the reaction between 1 (40 mg, 0.051
mmol) and ꢀ-caprolactone (12 mg, 0.11 mmol). Anal. Calcd for
C
26H40ClO2Y: Y, 17.5. Found: Y, 17.9. 1H NMR (C6D6): δ 3.35
(m, 2H), 2.40 (m, 2H), 2.09 (s, 30H), 1.22 (m, 2H), 0.95 (m,
4H). 13C NMR (C6D6): δ 185.7 (CO), 117.2 (C5Me5), 33.4 (CH2),
31.9 (CH2), 28.2 (CH2), 28.1 (CH2), 22.7 (CH2), 11.5 (C5Me5).
IR: 2904 s, 2866 s, 1688 vs, 1437 m, 1359 m, 1305 m, 1259 m,
1189 m, 1097 m, 1043 m, 803 w, 703 w cm-1
.
(C5Me5)2YCl(HMP A) (6). A white solid (68 mg, 93%) was
obtained from HMPA (23 mg, 0.12 mmol) and 1 (50 mg, 0.063
mmol). Anal. Calcd for C26H66ClN3OPY: Y, 15.0. Found: Y,
15.3. 1H NMR (C6D6): δ 2.18 (s, 30H), 2.18 (d, 18H, J ) 10
Hz). 13C NMR (C6D6): δ 117.0 (C5Me5), 37.4 (CH2, J C-P ) 20
Hz), 12.0 (C5Me5). IR: 2920 s, 2858 s, 1591 w, 1444 m, 1305 s,
1189 s, 1128 vs, 989 vs, 757 w cm-1
.
(C5Me5)2YCl(E-ca p r ola cta m ) (7). A white solid (68 mg,
94%) was isolated from ꢀ-caprolactam (16 mg, 0.14 mmol) and
1 (56 mg, 0.071 mmol) in 15 mL of benzene. Anal. Calcd for
C26H41ClNOY: Y, 17.5. Found: Y, 18.0. 1H NMR (C6D6): δ 8.67
(br, 1H, ∆ν1/2 ) 28 Hz), 2.30 (m, 2H), 2.08 (s, 30H), 1.90 (m,
2H), 1.17 (m, 2H), 1.08 (m, 2H), 0.88 (m, 2H). 13C NMR
(12) SMART Software Users Guide, Version 5.1; Bruker Analytical
X-ray Systems, Inc., Madison, WI, 1999.
(13) SAINT Software Users Guide, Version 6.0; Bruker Analytical
X-ray Systems, Inc., Madison, WI, 1999.
(14) Sheldrick, G. M. SADABS; Bruker Analytical X-ray Systems,
Inc.; Madison, WI, 1999.
(15) Sheldrick, G. M. SHELXTL Version 5.10; Bruker Analytical
X-ray Systems, Inc.; Madison, WI, 1999.
(11) Taylor, M. D.; Carter, C. D. J . Inorg. Nucl. Chem. 1962, 24,
387-391.
(16) International Tables for X-ray Crystallography; Kluwer Aca-
demic: Dordrecht, The Netherlands, 1992; Vol. C.

Lewis Base Coordination to an Electropositive Metal
Organometallics, Vol. 21, No. 9, 2002 1827
Ta ble 1. Su m m a r y of Cr ysta llogr a p h ic Da ta for (C5Me5)2YCl(L) (L ) THF (2), Ben zop h en on e (3), Meth yl
Meth a cr yla te (4), E-Ca p r ola cton e (5), Hexa m eth ylp h osp h or a m id e (6), E-Ca p r ola cta m (7),
1-Meth yl-2-p yr r olid in on e (8), N,N′-Dim eth ylp r op ylen eu r ea (9)) a n d (C5Me5)2YCl(E-ca p r ola cta m )2 (10)
3
4
5
6
7
8
9
10
formula
C
33H40Cl-
OY
C25H38Cl-
NO2Y
494.91
163(2)
C26H40Cl-
O2Y
508.94
158(2)
C26H48ClN3- C26H41Cl-
C25H39Cl-
NOY
493.93
173(2)
C26H42Cl-
N2OY
522.98
163(2)
monoclinic
P21/c
9.4446(4)
17.3668(7)
16.3932(7)
C32H52ClN2-
O2Y‚C6H11NO
734.27
163(2)
monoclinic
P21/c
18.3535(14)
16.4063(12)
12.7998(9)
OPY
574.00
158(2)
NOY
507.96
163(2)
fw
577.01
158
monoclinic
P21/c
13.0017(15) 16.4172(8)
10.7469(12) 14.9659(7)
21.515(2)
98.192(2)
2975.6(6)
4
temp (K)
cryst syst
space group
a (Å)
b (Å)
c (Å)
orthorhombic orthorhombic monoclinic
Pbca
orthorhombic monoclinic
Pbca
P21/c
Pbca
Pn
17.6071(10)
15.7221(9)
18.8986(10)
90
5231.5(5)
8
1.292
2.350
0.0560
10.1251(8)
17.4491(8)
10.4464(7)
9.7393(6)
12.3895(8)
93.7970(10) 91.8030(10) 92.1630(10)
1257.75(14) 2687.52(19) 3851.4(5)
19.0600(16) 15.5408(7)
15.3784(13) 19.1854(9)
20.5069(10)
90
5038.5(4)
8
1.305
2.438
â (deg)
V (Å3)
Z
95.996(2)
2951.6(4)
4
90
5202.6(4)
8
2
4
4
Fcalcd (Mg/m3) 1.288
1.292
2.142
0.0629
1.297
2.361
0.0583
1.304
2.440
0.0449
1.293
2.288
0.0266
1.266
1.621
0.0505
µ (mm-1
)
2.072
0.0546
final R1
0.0335
(I > 2σ(I))
final wR2
(all data)
0.1406
0.0849
0.1466
0.1829
0.1362
0.1094
0.0675
0.1428
(C5Me5)2YCl(E-ca p r ola cta m ) (7). Hydrogen atom H(1) was
located from a difference Fourier map and refined (x, y, z, and
U
iso). The remaining hydrogen atoms were included using a
riding model.
(C5Me5)2YCl(1-m eth yl-2-p yr r olid in on e) (8). The absolute
structure was assigned by refinement of the Flack parameter.16
(C5Me5)2YCl(DMP U) (9). Hydrogen atoms were located
from a difference Fourier map and refined (x, y, z, and Uiso).
(C5Me5)2YCl(E-ca p r ola cta m )2 (10). There was one solvent
molecule of ꢀ-caprolactam present per formula unit. The
solvent exhibited high thermal motion, but a suitable disorder
model could not be established. The largest difference peak
was 1.7 e/Å3. The solvent was included using SAME and EADP
constraints in SHELXTL.15
Resu lts
Syn t h esis of (C5Me5)2YCl(L) Com p lexes. (C5-
Me5)2Y(µ-Cl)Y(C5Me5)2Cl (1) reacts readily with oxygen
donor Lewis bases to form the monometallic adducts (C5-
Me5)2YCl(L) (L ) THF1 (2), benzophenone (3), methyl
methacrylate (MMA; 4), ꢀ-caprolactone (5), hexameth-
ylphosphoramide (HMPA; 6), ꢀ-caprolactam (7), 1-meth-
yl-2-pyrrolidinone (MPU; 8), N,N′-dimethylpropyle-
neurea (DMPU; 9)), according to eq 1. In each case the
F igu r e 1. Molecular structure of (C5Me5)2YCl(benzo-
phenone) (3) with probability ellipsoids drawn at the 50%
level.
Ta ble 2. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for Com p lexes (C5Me5)2YCl(L)1 (2-9) a n d
(C5Me5)2YCl(E-ca p r ola cta m )2 (10)
Ln-
(C5Me5
ring
(ring
diff in Y-O
length,
centroid)-
metal-(ring
centroid)
com-
complex N -
plex centroid)
Ln-O
complex N - 1
Ln-Cl
(C Me ) Y(µ-Cl)Y(C Me ) Cl
+ 2L f
5
5 2
5
5 2
21
2.382
2.379
2.376
2.384
2.386
2.382
2.367
2.361
2.401
2.422
2.377
2.379
2.382
2.384
2.396
2.394
2.425
2.435
136.2
138.0
137.2
137.4
134.0
137.0
136.4
136.9
138.3
2.410(7)
2.579(3)
1
3
2.312(2)
2.306(2)
2.263(3)
2.255(3)
2.249(3)
2.233(3)
2.218(1)
0.098
0.006
0.043
0.008
0.006
0.016
0.015
2.576(1)
2.590(1)
2.565(1)
2.588(1)
2.606(1)
2.600(1)
2.6022(4)
2.758(1)
(C Me ) YCl(L)
2
(1)
5
5 2
2-9
4
L ) THF (2), benzophenone (3), MMA (4),
ꢀ-caprolactone (5), HMPA (6), ꢀ-caprolactam (7),
5
6
MPU (8), DMPU (9)
7
reaction can be carried out with a stoichiometric amount
of L and the yield is high (>80%). Complexes 2 and 5-9
are isolable as white crystalline solids, (C5Me5)2YCl-
(MMA) (4) is yellow, and (C5Me5)2YCl(benzophenone)
(3) is a dark red solid due to a broad absorption with a
maximum at λ ) 457 nm with ꢀ ) 1500 M-1 cm-1. Color
changes due to benzophenone coordination have been
observed previously with organoaluminum complexes.17
8
9
10
2.346(2)
2.348(2)
Str u ctu r a l Com p a r ison s on th e (C5Me5)2YCl(L)
Com p lexes. Table 2 presents structural information on
3-9, which are shown in Figures 1-8, as well as (C5-
Me5)2YCl(THF) (2).1 In all of these complexes, the Lewis
(17) Power, M. B.; Bott, S. G.; Atwood, J . J .; Barron, A. R. J . Am.
Chem. Soc. 1990, 112, 3446-3451.

1828 Organometallics, Vol. 21, No. 9, 2002
Evans et al.
F igu r e 2. Molecular structure of (C5Me5)2YCl(MMA) (4)
with probability ellipsoids drawn at the 50% level.
F igu r e 5. Molecular structure of (C5Me5)2YCl(ꢀ-caprolac-
tam) (7) with probability ellipsoids drawn at the 50% level.
F igu r e 3. Molecular structure of (C5Me5)2YCl(ꢀ-caprolac-
tone) (5) with probability ellipsoids drawn at the 30% level.
Disorder in the caprolactone and the C(11)-C(15) ring is
not shown.
F igu r e 6. Molecular structure of (C5Me5)2YCl(MPU) (8)
with probability ellipsoids drawn at the 50% level.
F igu r e 4. Molecular structure of (C5Me5)2YCl(HMPA) (6)
with probability ellipsoids drawn at the 50% level.
F igu r e 7. Molecular structure of (C5Me5)2YCl(DMPU) (9)
with probability ellipsoids drawn at the 50% level.
base coordinates to yttrium via a single oxygen atom.
Hence, in 5 (Figure 3), ꢀ-caprolactone coordinates as an
η1 ligand through the carbonyl oxygen despite the
presence of an adjacent oxygen donor atom next to the
carbonyl oxygen. This is consistent with the only other
ꢀ-caprolactone/yttrium complex reported in the litera-
ture: mer-YCl3(ꢀ-caprolactone)3.5 The η1 binding mode
of ꢀ-caprolactam in 7 (Figure 5) is similar to that found
in [Pr(C6H11NO)7]3+ 7 [Pr(C6H11NO)6Cl]2+ 8
, , and [Eu-

Lewis Base Coordination to an Electropositive Metal
Organometallics, Vol. 21, No. 9, 2002 1829
in solution, this lattice caprolactam exchanges with the
solvated caprolactam.
As shown in Table 2, complexes 2-9 have similar
structural parameters in the metallocene part of the
molecules. Hence, the (C5Me5 ring centroid)-yttrium-
(C5Me5 ring centroid) angles fall in a narrow 134-138°
range, as do the Y-(C5Me5) ring centroid distances, at
2.36-2.43 Å. The Y-Cl distances, which sometimes can
be more variable in a series,1 are also quite regular:
2.565(1)-2.606(1) Å.
In contrast to the similar structural parameters of the
(C5Me5)2YCl parts of these molecules, the Y-O dis-
tances are more variable with distances ranging from
2.218(1) to 2.410(7) Å. Since this range is not extremely
large and since some of the differences between the
ligands are within the error limits, care must be taken
in using these data. However, these results do show that
these oxygen donors approach yttrium at slightly dif-
ferent distances, despite the relatively uniform (C5Me5)2-
YCl components.
The longest of the Y-O distances, 2.410(7) Å, is found
in the THF adduct 2.1 Since THF is a common solvent
in organometallic yttrium chemistry, this provides a
good reference point. The next longest bond lengths are
the 2.312(2) Å distance found in the benzophenone
adduct 3 and the 2.306(2) Å distance in the methyl
methacrylate complex 4. Structural data on lanthanide
coordination complexes of MMA have not been reported
to our knowledge, but the yellow terbium benzophenone
complex Tb[N(SiMe3)2]3(benzophenone) has a Tb-O
distance of 2.305(9) Å.19 This distance is comparable to
the Y-O distance in 3, since terbium is approximately
0.02 Å larger than yttrium.20 An yttrium benzophenone
complex has also been synthesized, Y[N(SiMe3)2]3-
(benzophenone),19 but it was not structurally character-
ized and, unlike 3, did not display absorption in the
visible region.
F igu r e 8. Molecular structure of (C5Me5)2YCl(ꢀ-caprolac-
tam)2 (10) with probability ellipsoids drawn at the 50%
level.
(C6H11NO)4Cl2]+.8 The other substituted “lactam type”
molecules, the 1-methyl-2-pyrrolidinone adduct 8 (Fig-
ure 6) and the N,N′-dimethylpropyleneurea adduct 9
(Figure 7), also adopt η1 binding modes. Methyl meth-
acrylate coordinates to the metal center in 4 only
through the carbonyl oxygen. The other oxygen and the
adjacent vinyl group point away from the metal (eq 2,
Figure 2). The η1 binding mode of HMPA (Figure 4) is
typical of that seen in other trivalent lanthanide/HMPA
complexes such as [TmI2(HMPA)4]I(pyridine)5.18
ꢀ-Caprolactone, the precursor to a biodegradable
polylactone,21 forms the next smallest Y-O bond in this
series, which is 2.263(3) Å in 5. This distance is similar
to the 2.269(2)-2.296(3) Å range of distances found in
Since each complex has a single Y-O(donor atom)
bond length, this series can be used to compare the first
approach of these ligands to a trivalent Lewis acidic
metal center in a metallocene coordination environment.
It might be assumed that coordination through a single
donor atom would necessarily occur because the (C5-
Me5)2YCl moiety only has room for one more ligand.
However, isolation of (C5Me5)2YCl(C6H11NO)2 (10) (Fig-
ure 8) shows that this is not the case. Reaction of 1 with
excess ꢀ-caprolactam gives a bis adduct (eq 3). Complex
5
mer-YCl3(ꢀ-caprolactone)3 despite the difference in
formal coordination number, 6 vs 8 in 5.
The next smallest Y-O distance of 2.255(3) Å was
found with hexamethylphosphoramide (HMPA), which
was examined since it has been utilized extensively as
a strong donor to increase the reduction potential of Sm-
(II) in SmI2.22-25 The 2.255(3) Å Y-O(HMPA) distance
in 6 is reasonable compared to trivalent Tm-O(HMPA)
distances of 2.189(4) Å in [TmI2(HMPA)4]I(pyridine)5
and 2.198(6) Å in {[TmI(HMPA)4(pyridine)][I]2},18 when
radial sizes are taken into consideration.20
ꢀ-Caprolactam, the precursor to nylon-6,26 forms the
next Y-O bond in the decreasing order, the 2.249(3) Å
(19) Allen, M.; Aspinall, H. C.; Moore, S. R.; Hursthouse, M. B.;
Karvalov, A. I. Polyhedron 1992, 11, 409-413.
(20) Shannon, R. D. Acta Crystallogr. 1976, A32, 751-767.
(21) Yasuda, H.; Tamai, H. Prog. Polym. Sci. 1993, 18, 1097-1139
and references therein.
(22) Otsubo, K.; Inanaga, J .; Yamaguchi, M. Tetrahedron Lett. 1986,
27, 5763-5764.
(23) Curran, D. P.; Fevig, T. L.; Elliott, R. L. J . Am. Chem. Soc.
1988, 110, 5064-5067.
10 crystallizes with one molecule of ꢀ-caprolactam in the
lattice. The NMR spectrum of this system indicates that
(24) Molander, G. A.; Harring, L. S. J . Org. Chem. 1990, 55, 6171-
6176.
(18) Evans, W. J .; Broomhall-Dillard, R. N. R.; Ziller, J . W.
Polyhedron 1998, 17, 3361-3370.
(25) Shiue, J . S.; Lin, C. C.; Fang, J . M. Tetrahedron Lett. 1993, 34,
335-338.

1830 Organometallics, Vol. 21, No. 9, 2002
Evans et al.
Y-O bond in 7. This distance is shorter than the
distances found in cationic [Pr(C6H11NO)6Cl]2+ and [Eu-
(C6H11NO)4Cl2]+, 2.418(3) and 2.315(18) Å, respectively,8
when metal size is taken into account.20 The 2.58(6) Å
NH‚‚‚Cl and 3.328(5) Å N‚‚‚Cl distances found in 7 are
appropriate for hydrogen bonding compared to analo-
gous distances in, for example, ꢀ-caprolactam hydro-
chloride27 and the tris(pyrazolyl)hydroborate {[HB(3-
ButpzH)3]Cl}[AlCl4].28
shorter Y-O bond (eq 4). The reactions were monitored
(C5Me5)2YCl(L) + L′ f 2(C5Me5)2YCl(L′) + L
(4)
1
by H NMR spectroscopy in C6D6 to determine if the
ligand with the shorter Y-O bond would displace the
ligand with a longer Y-O bond.
Addition of benzophenone to (C5Me5)2YCl(THF) (2)
1
instantaneously forms a dark red solution of 3. The H
NMR spectrum indicated complete displacement of THF
by benzophenone to form (C5Me5)2YCl(benzophenone)
(3): the 2.20 ppm C5Me5 shift of 2 is absent, the 2.04
ppm shift of 3 is the only peak in the C5Me5 region, and
free THF is observed in the spectrum. Additional
reactions were initially done with 2 to confirm the
generality of this result: THF was also displaced by
MMA and ꢀ-caprolactone to form (C5Me5)2YCl(MMA) (4),
and (C5Me5)2YCl(ꢀ-caprolactone) (5), respectively. All of
these displacements formed products with shorter Y-O
bonds.
The shortest of the Y-O distances observed were from
1-methyl-2-pyrrolidinone (MPU), 2.233(3) Å in 8, and
N,N′-dimethylpropyleneurea (DMPU), 2.218(1) Å in 9.
MPU was included to examine the effect on binding of
replacing an N-H functionality in 7 with N-Me. DMPU
has been explored in lanthanide chemistry as a possible
replacement for carcinogenic HMPA in Sm(II) reduction
chemistry.29,30
The structure of (C5Me5)2YCl(ꢀ-caprolactam)2 (10) is
similar to those of 2-9 in the metallocene region but
different in the other distances. The 138.3° centroid-
metal-centroid angle and 2.425 and 2.435 Å Y-(C5Me5)
ring centroid distances in 10 are near the ranges found
in 2-9, 134-138° and 2.36-2.43 Å, respectively (Table
2). The most striking difference in 10 is the 2.7576(9) Å
Y-Cl distance, which is substantially larger than the
2.565(1)-2.606(1) Å distances in 2-9. This unusually
long distance is probably due to hydrogen bonding, since
the chloride anion is located between two ꢀ-caprolactam
molecules at distances appropriate for this type of
interaction. The N‚‚‚Cl distances of 3.161(3) and 3.174-
(3) Å are shorter than the analogous distance in 7,
3.328(5) Å. The 2.348(2) and 2.346(2) Å Y-O(ꢀ-capro-
lactam) distances are also longer that the 2.249(3) Å
distance in 7 (nine-coordinate yttrium is 0.056 Å larger
than eight-coordinate yttrium).20
However, when MMA is added to (C5Me5)2YCl(benzo-
phenone) (3), the red color of 3 persists. Since the C5-
Me5 chemical shifts of 3 and (C5Me5)2YCl(MMA) (4) are
similar, 2.04 and 2.02 ppm, respectively, the reaction
was monitored by observing MMA resonances. No
displacement was observed: only free MMA was found
in the 1H NMR spectrum. In this case, the reverse
reaction of yellow (C5Me5)2YCl(MMA) (4) with benzo-
phenone occurs quickly to form a dark red solution
1
containing only 3 and free MMA by H NMR spectros-
copy. These solution data suggest a preference for
benzophenone coordination over MMA, although (C5-
Me5)2YCl(MMA) has a numerically smaller Y-O value
(2.306(2) Å) than (C5Me5)2YCl(benzophenone (2.312(2)
Å).
When (C5Me5)2YCl(benzophenone) is reacted with
ꢀ-caprolactone, the dark red color dissipates instanta-
neously to form a colorless solution. By 1H NMR
spectroscopy, benzophenone is displaced by ꢀ-caprolac-
tone to form (C5Me5)2YCl(ꢀ-caprolactone) (5). Similarly,
when (C5Me5)2YCl(MMA) (4) is reacted with ꢀ-capro-
lactone, the yellow color immediately dissipates and only
In summary, numerical values of the Y-O distances
in 2-9 decrease in the order THF, benzophenone, MMA,
ꢀ-caprolactone, HMPA, ꢀ-caprolactam, MPU and DMPU.
The THF-benzophenone difference is the largest at
0.098 Å, and the benzophenone-MMA and HMPA-ꢀ-
caprolactam differences (0.006 Å) are the smallest
(Table 2). The other differences are in the range 0.008-
0.043 Å.
1
5 is observed by H NMR.
Reactions of HMPA with (C5Me5)2YCl(ꢀ-caprolactone)
(5), ꢀ-caprolactam with (C5Me5)2YCl(HMPA) (6), and
DMPU with (C5Me5)2YCl(MPU) (8) occur quickly and
For the complexes of ligands containing carbonyl
moieties, 3-5 and 7-9, the Y-O distances were com-
pared with Y-O-C angles. As shown by the following
sequence, which decreases in order of Y-O distance,
there is no correlation: 3, benzophenone, 169.8(3)°; 4,
MMA, 170.44(17)°; 5, ꢀ-caprolactone, 156.5(3)°; 7, ꢀ-
caprolactam, 146.5(3)°; 8, MPA, 171.6(3)°; 9, DMPU,
163.58(11)°.
1
are complete as soon as the H NMR spectrum can be
obtained. In each case ligand displacement occurs to
form a complex with a shorter bond to yttrium. How-
ever, the trend between Y-O metal distance and
solution displacement is not observed with (C5Me5)2YCl-
(ꢀ-caprolactam) (7), as neither MPU nor DMPU displace
ꢀ-caprolactam from 7. The extra stability of 7 is consis-
tent with its hydrogen-bonded structure.
Liga n d Disp la cem en t Rea ction s of (C5Me5)2YCl-
(L) Com p lexes. To determine if the solid-state Y-O
bond lengths correlated with solution behavior, each (C5-
Me5)2YCl(L) yttrium adduct was reacted with an equi-
molar amount of the neutral donor that had the next
IR a n d 13C NMR Sp ectr oscop y of th e Ca r bon yl-
Con ta in in g Com p lexes. The ∆νCO stretching frequen-
cies and 13C NMR shifts of the carbonyl carbon in
compounds 3-5 and 7-9 were examined to see if the
shift observed from the free ligands correlated with the
Y-O bond distances and ligand displacement data. As
shown in Table 3, these data do not correlate exactly
with the Y-O bond lengths or the displacement reactiv-
ity. The IR data show that the complexes with the
shortest Y-O distances, 8 and 9, have the largest
differences in ∆νCO between complexed and free ligand,
(26) Vasiliu-Oprea, C.; Dan, F. J . Appl. Polym. Sci. 1996, 62, 1517-
1527 and references therein.
(27) Winkler, F. K.; Dunitz, J . D. Acta Crystallogr. 1975, B31, 273-
275.
(28) Parkin, G.; Looney, A.; Rheingold, A. Inorg. Chem. 1991, 30,
3099-3101.
(29) Curran, D. P.; Wolin, R. L. Synlett 1991, 5, 317-318.
(30) Flowers, R. A.; Caracoti, A. F.; Sealy, J . R.; Shabangi, J . Recent
Res. Dev. Org. Chem. 1999, 3, 141-149.

Lewis Base Coordination to an Electropositive Metal
Organometallics, Vol. 21, No. 9, 2002 1831
Ta ble 3. Ca r bon yl Str etch in g F r equ en cies (cm -1
a n d 13C NMR Sh ifts (p p m ) for th e F r ee Liga n d s
)
within the experimental error limits, it was unclear if
these data would be useful in predicting chemical
reactivity. Surprisingly, there is a rough correlation
between the solution displacement chemistry and the
solid-state data.
a n d (C5Me5)2YCl(L) Com p lexes 3-5 a n d 7-9, Listed
in Or d er of Decr ea sin g Y-O Dista n ce
IR, cm-1
NMR, ppm
com-
com-
Hence, bases that form complexes with shorter Y-O
distances will displace bases that form complexes with
longer Y-O distances, with two exceptions. The first is
the benzophenone/MMA pair, which differ in measured
Y-O distances by only 0.006 Å, a difference too small
upon which to make a prediction. The other exception
is that the ꢀ-caprolactam in complex 7 cannot be easily
displaced by bases which form shorter Y-O bonds. This
too can be rationalized, in this case, by the fact that
ꢀ-caprolactam forms a hydrogen bond to the chloride,
which provides extra stabilization. Hence, this (C5-
Me5)2Y(µ-Cl)Y(C5Me5)2Cl/(C5Me5)2YCl(L) system can be
used not only to examine Y-O bonding but also to
examine the influence of hydrogen bonding.
This study also showed that the (C5Me5)2YCl coordi-
nation envirnoment is flexible enough to accommodate
two base adducts, as shown in (C5Me5)2YCl(ꢀ-caprolac-
tam)2 (10). In this case, two N-H‚‚‚Cl hydrogen bonds
resulted in a rather long Y-Cl bond. Hence, hydrogen
bonding of this type can be used to destabilize certain
metal-ligand bonds in metallocene complexes of this
type.
Attempts to correlate ∆νCO carbonyl stretching fre-
quencies or 13C NMR shifts of carbonyl carbons with
the displacement reactivity were not as successful.
Hence, the bond length data appear to be the best
available predictors of displacement reactivity. As long
as it is realized that small differences in Y-O bond
lengths may not give predictable results and additional
sources of stabilization can interfere with correlations,
the length of the Y-donor atom interactions in the (C5-
Me5)2YCl(L) system can be useful in estimating the
relative ability of different ligands to coordinate. In any
case, it is a good system to evaluate ligand binding
modes in a metallocene environment.
com-
plex
free plexed
free plexed
ligand (L)
L
L
diff
L
L
diff
3
4
5
7
8
benzophenone
methyl methacrylate 1748 1661
1684 1614
70 195.7 206.6 10.9
87 167.2 175.5 8.3
84 175.0 185.7 10.7
ꢀ-caprolactone
ꢀ-caprolactam
1-methyl-2-
pyrrolidinone
N,N′-dimethyl-
propyleneurea
1772 1688
1713 1630
1738 1645
83 177.0 181.9
93 173.7 176.6
4.9
2.9
9
1687 1576 111 156.3 157.3
1
but the other ligands do not follow a regular order and
the differences in stretching frequencies are in a small
range. The 13C NMR data show that complexes 8 and 9
have the smallest differences between free and com-
plexed ligand and that the carbonyl complex with the
longest Y-O bond, 3, has the largest difference, but the
value for the MMA complex 4 is out of order.
Discu ssion
Although a few oxygen-donor complexes of yttrium
permethylmetallocenes of general formula (C5Me5)2YX-
(L) (X ) monoanionic ligand, L ) oxygen donor ligand)
have been structurally characterized prior to this study,
e.g. (C5Me5)2YCl(THF),1 (C5Me5)2Y(CH3)(THF),31 (C5-
Me5)2Y(C6H3Me2-2,6)CdN(C6H3Me2-3,5)(THF),32 and (C5-
Me5)2Y(CH2Ph)(THF),33 all involve only L ) THF.
Adducts of reactive polymerizable substrates such as
ꢀ-caprolactone and methyl methacrylate would be more
difficult to isolate, since these substrates can react with
yttrium alkyl bonds.34,35 The chloride complex (C5-
Me5)2Y(µ-Cl)Y(C5Me5)2Cl (1) appears to be a good system
with which to evaluate the coordination chemistry of
Lewis bases, since polymerizable substrates do not react
under ambient conditions. A variety of adducts can be
synthesized in high yield, and suitable crystals can be
obtained such that metrical data on coordination chem-
istry can be obtained.
Con clu sion
Although this survey has focused on polymerizable
substrates and typical lanthanide oxygen donor base
adducts, the results suggest that 1 could be used more
generally as a route to make crystalline adducts of a
variety of oxygen donor molecules. For example, if the
absolute configuration of an organic ether or ester was
needed and the molecule did not crystallize on its own,
adduct formation with 1 could provide the needed
structural information. This could prove to be a valuable
aspect of this system.
(C5Me5)2Y(µ-Cl)Y(C5Me5)2Cl provides an effective
means for the comparative evaluation of ligand donor
ability in typical organoyttrium and organolanthanide
complexes. It is a readily synthesized, diamagnetic
complex which forms crystalline adducts, (C5Me5)2YCl-
(L), in high yield. The Y-O bond distances correlate
with the substitution reactivity of this class of complexes
and allow examination of hydrogen-bonding effects. This
approach should be readily extendable to other ligands
and substrates to provide data on their binding and
structure as well as a larger body of data which allows
direct comparisons in a metallocene system.
In each case reported here, the oxygen-containing
adduct binds only through a single oxygen such that the
Y-O bond lengths can be compared. Since the Y-O
distances span only a short range and since several are
Ack n ow led gm en t. For the support of this research,
we thank the Division of Chemical Sciences of the Office
of Basic Energy Sciences of the Department of Energy.
(31) Den Haan, K. H.; de Boer, J . L.; Teuben, J . H.; Smeets, W. J .;
Spek, A. L. J . Organomet. Chem. 1987, 327, 31-38.
(32) Mandel, A.; Magull, J . Z. Anorg. Allg. Chem. 1996, 622, 1913-
1919.
(33) Schumann, H.; Rosenthal, E. C. E.; Kociok-Ko¨ln, G. Molander,
G. A.; Winterfield, J . J . Organomet. Chem. 1995, 496, 233-240.
(34) Yasuda, H.; Ihara, E.; Takemoto, Y.; Yamashita, M. Macro-
molecules 1996, 29, 1798-1806.
(35) Yasuda, H.; Yamamoto, H.; Yokota, K.; Miyake, S.; Nakamura,
A. J . Am. Chem. Soc. 1992, 114, 4908-4910.
Su p p or tin g In for m a tion Ava ila ble: Tables giving X-ray
crystallographic data for complexes 3-10; these data are also
available in CIF format. This material is available free of
charge via the Internet at http://pubs.acs.org.
OM010908D