Lanthanide complexes of an oxazoline-phenoxide hybrid chelate

Article abstract of DOI:10.1139/v05-069

The synthesis of 2-(2′-hydroxy-3′-allylphenyl)-4,4-dimethyl-2- oxazoline, H-Allox (3), and lanthanide tris chelate complexes, mer-Ln(Allox)₃ (Ln = La (4), Ce (5), Sm (6), Er (7), and Y (8)), derived from it are reported. A six-coordinate mer ge

Full text of DOI:10.1139/v05-069

449
Lanthanide complexes of an oxazoline-phenoxide
hybrid chelate
David J. Berg, Chuanjian Zhou, Tosha Barclay, Xuening Fei, Shengyu Feng,
Kevin A. Ogilvie, Robert A. Gossage, Brendan Twamley, and Mark Wood
Abstract: The synthesis of 2-(2′-hydroxy-3′-allylphenyl)-4,4-dimethyl-2-oxazoline, H-Allox (3), and lanthanide tris
chelate complexes, mer-Ln(Allox)3 (Ln = La (4), Ce (5), Sm (6), Er (7), and Y (8)), derived from it are reported. A
six-coordinate mer geometry without alkene coordination was confirmed in the solid state by X-ray crystallography for
5 and 7. Variable-temperature NMR experiments suggested that this is the most stable isomer in solution as well, al-
though the inequivalent ligand environments undergo rapid averaging at room temperature for all five complexes. A
mechanistic investigation indicated that this fluxional process is an intramolecular six-coordinate rearrangement, but it
was not possible to distinguish between a Bailar (trigonal) or Rây–Dutt (rhombic) twist. Kinetic parameters for the
fluxional process were determined by line shape analysis for 8 yielding ∆H‡ = 24 2 kJ mol–1 and ∆S‡ = –99
10 J mol–1 K–1. The structural and dynamic features of 4–8 were compared with the related In, Ga, and Al tris(2-
oxazolylphenoxides).
Key words: oxazoline, phenoxide, lanthanide, Group 3, Rây–Dutt twist, Bailar twist, Eyring plot, X-ray crystallography,
mer isomer, variable-temperature NMR, line shape analysis, dynamic NMR, paramagnetic NMR, bidentate ligands.
Résumé : On rapporte la synthèse de la 2-(2′-hydroxy-3′-allylphényl)-4,4-diméthyl-2-oxazoline, H-Allox (3) et de
complexes de lanthanide trischélates, mer-Ln(Allox)3 (Ln = La (4), Ce (5), Sm (6), Er (7) et Y (8)) qui en dérivent.
Pour les complexes 5 et 7, les données cristallographiques ont permis de confirmer que, à l’état solide, ces complexes
adoptent une géométrie mer six coordinnées sans coordination d’alcène. Les données d’expériences de RMN à tempéra-
ture variable suggèrent que, en solution, cet isomère est aussi le plus stable même si les environnements inéquivalents
des ligands de chacun de ces cinq complexes sous soumis à une moyenne rapide. Une investigation mécanistique in-
dique que ce processus fluxionnel correspond à un réarrangement intramoléculaire à six coordonnées, mais il n’a pas
été possible de faire une distinction entre une torsion de Bailar (trigonale) ou de Rây–Dutt (rhombique). On a déter-
miné les paramètres cinétiques du processus fluxionnel par une analyse de la forme des bandes du composé 8 et on en
a déduit des valeurs de ∆H‡ = 24 2 kJ mol–1 et ∆S‡ = –99 10 J mol–1 K–1. Les caractéristiques structurales et dy-
namiques des composés 4–8 avec celles des composés apparentés des tris(2-oxazolylphénolates) d’indium, de gallium
et d’aluminium.
Mots clés : oxazoline, phénolate, lanthanide, Groupe 3, torsion de Rây–Dutt, torsion de Bailar, courbe de Eyring,
cristallographie par diffraction des rayons X, isomère mer, RMN à température variable, analyse de la forme des
bandes, RMN dynamique, RMN paramagnétique, ligands bidentates.
[Traduit par la Rédaction] Berg et al. 459
Introduction
type is found in nature as part of microbial metal transport
agents, especially the Fe(III) transporting siderophores (1,
2). Second, asymmetric ligands of this kind are easily pre-
pared from readily available chiral β-aminoalcohols and the
resulting chiral metal complexes have proven useful as
The transition metal and Group 13 chemistry of chelating
2-(2′-hydroxyphenyl)-2-oxazoline ligands (H-I) has been ex-
tensively studied for two principal reasons. First, this chelate
Received 14 January 2005. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 10 May 2005.
D.J. Berg1 and M. Wood. Department of Chemistry, University of Victoria, P.O. Box 3065, Victoria, BC V8W 3V6, Canada.
C. Zhou. School of Materials Science and Engineering, Shandong University, 73 Jing Shi Lu Ji’nan City, Shandong Province, P.R.
China 250061.
T. Barclay. Chemistry Department, North Harris College, 2700 W.W. Thorne Blvd. Houston, TX 77073-3499, USA.
X. Fei. Tianjin Institute of Urban Construction, 26 Jinjing Road, Xiqing District, Tianjin, P.R. China 300384.
S. Feng. Department of Chemistry and Chemical Engineering, Shandong University, 27 Shandu Nanlu, Ji’nan City, Shandong
Province, P.R. China 250100.
K.A. Ogilvie and R.A. Gossage. The Chester Woodleigh Small Laboratory of Organic Chemistry, Department of Chemistry,
Acadia University, Wolfville, NS B4P 2R6, Canada.
B. Twamley. University Research Office, 109 Morrill Hall, University of Idaho, Moscow, ID 83844-3010, USA.
1Corresponding author (e-mail: djberg@uvic.ca).
Can. J. Chem. 83: 449–459 (2005)
doi: 10.1139/V05-069
© 2005 NRC Canada

450
Can. J. Chem. Vol. 83, 2005
asymmetric oxidation catalysts (2). Despite the interest in 2-
(2′-hydroxyphenyl)-2-oxazolines, and oxazolines in general,
as ligands in transition metal chemistry, the Group 3 and
lanthanide chemistry of such complexes has not been very
well explored. To date, the limited oxazoline coordination
chemistry that has been disclosed with these elements has
been almost exclusively restricted to the neutral methylene-
bis(oxazolines) (II, box) (3) or their close relatives contain-
ing additional donors: pyridine-bis(oxazolines) (III, pybox)
(3, 4), 2,2′-bipyridine-bis(oxazolines) (IV) (3), and dibenzofuran-
bis(oxazolines) (V) (3). This is remarkable considering the
success of these ligands in catalytically active complexes of
the d-block metals (2). Indeed, even though little work has
been done so far, enantiomerically pure II–V have proven
successful in Group 3 and lanthanide chemistry for asym-
metric hetero-Diels–Alder (3), glyoxylate-ene (4a), Mukayama–
Michael (4b), and annulation and addition reactions of
allenylsilanes with aldehydes (4c).
I
O
N
O
O
O
O
O
N
N
N
N
N
N
N
R
R
II
III
R
R
O
O
O
O
O
N
N
N
N
IV
V
R
R
R
R
Recently, examples of lanthanide complexes with anionic
oxazoline-containing ligands (I, VI, and VII) have been re-
ported (5–7). The tetrakis complex of 2-(2′-hydroxyphenyl)-
4,4-dimethyl-2-oxazoline (I) with Ce4+ was prepared as a
possible precursor to CeO2 films by thermal decomposition.
Lanthanide complexes of VI proved effective as asymmetric
catalysts for the 1,3-dipolar cycloaddition of nitrones to
alkenes, although the structure of the actual catalyst was not
established (being formed in situ from Yb(OTf)3 and VI)
(6). The deprotonated box ligand VII has been used to pre-
pare Ln[VII][NR2]2 (Ln = Y and R = SiHMe2; Ln = La, Nd,
Sm, Lu, and R = SiMe3) and Ln[VII][CH(SiMe3)2]2 (7)
complexes that serve as hydroamination catalysts.
In this paper, we report the synthesis, solid-state structure,
and solution behaviour of lanthanide complexes containing
deprotonated 2-(2′-hydroxy-3′-allylphenyl)-4,4-dimethyl-2-
oxazoline (Allox). The goal of this work is to understand the
structural chemistry of tris(oxazoline-phenoxide) complexes
of the lanthanides in solid state and solution. Although the
present complexes are achiral, their solution behaviour
should provide useful information about the suitability of
enantiomerically pure variants as asymmetric Lewis acid
catalysts. Besides providing some additional steric bulk, the
allyl substituent was deliberately incorporated to assess
whether coordination of the tethered alkene can occur. We
have previously observed rare examples of alkene coordina-
tion with lanthanide centres for a phenoxide ligand bearing a
tethered allyl group so it was of interest to see if this also
occurred with Allox.2
O
O
-
N
N
O O
- -
O
O
R
R
N
N
VI
VII
R
R
Schlenk techniques. Tetrahydrofuran (THF), hexane, and to-
luene were dried by distillation from sodium benzophenone
ketyl under argon immediately prior to use. The appropriate
lanthanide tris(silylamides), Ln[N(SiMe3)2]3, were prepared
according to literature procedures (8). All NMR spectra (1H,
13C) including variable-temperature experiments were recorded
on a Bruker AMX-360 MHz spectrometer and referenced to
residual solvent resonances. All deuterated solvents were
dried over activated 4 Å molecular sieves and spectra were
recorded using 5 mm tubes fitted with a Teflon® valve
(Brunfeldt). Variable-temperature NMR spectra and plots of
δ vs. 1/T for paramagnetic complexes 5–7 are included in the
Supplementary material,3 while a representative high-
temperature spectrum is included in the experimental data
for each complex given below. Melting points were recorded
using a Büchi melting point apparatus and are not corrected.
Elemental analyses were performed by Canadian Micro-
analytical, Delta, British Columbia. Mass spectra were re-
corded on a Kratos Concept H spectrometer using electron
impact (70 eV) ionization. The molecular ion simulation for
complex 6 is shown in Fig. 1; simulations for 4–7 are in-
cluded in the Supplementary material.3
Experimental
General procedures
All manipulations were carried out under an argon or ni-
trogen atmosphere, with the rigorous exclusion of oxygen
and water, using standard glovebox (Braun MB150-GII) or
2-(2′-Hydroxyphenyl)-4,4-dimethyl-2-oxazoline (1)
2-Cyanophenol (10.0 g, 0.0840 mol), Yb(CF3SO3)3
(0.52 g, 1.0 mol%), and 2-amino-2-methylpropanol (11.22 g,
2 D.J. Berg, T. Barclay, and X. Fei. Manuscript in preparation. Presented in part as contribution #764 at the 85th Canadian Society for Chem-
istry Conference, Vancouver, British Columbia, 1–6 June 2002.
3 Supplementary data for this article are available on the Web site or may be purchased from the Depository of Unpublished Data, Document
Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0S2, Canada. DUD 3656. For more information on obtaining mate-
rial refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 260761 and 260762 contain the crystallographic data for this manu-
script. These data can be obtained, free of charge, via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2005 NRC Canada

Berg et al.
451
Fig. 1. (a) Experimental and (b) calculated isotopic patterns for the molecular ion of Sm(Allox)3 (6).
(a)
(b)
0.126 mol) were suspended in dry 1,2-dichlorochloro-
benzene (150 mL) in a Kontes flask under argon. The flask
was sealed and the reaction mixture was heated to 170 °C
for 24 h; all solids dissolved upon heating. After cooling, the
flask was opened and the contents were poured onto ice
(200 g). Dichloromethane (200 mL) was added and the two-
phase mixture was separated. The organic layer was washed
repeatedly with water, dried over anhydr. MgSO4, filtered,
and the solvent removed from the filtrate under vacuum to
afford 2-(2′-hydroxyphenyl)-4,4-dimethyl-2-oxazoline (9) as
1H, CH=CHcis, J = 11.1 Hz), 4.46 (d, 2H, OCH2CH=CH2,
J = 6.6 Hz), 3.98 (s, 2H, CH2), 1.32 (s, 6H, CH3). HR-MS
(EI) calcd.: M+ 231.1259; found: 231.1271.
2-(2′-Hydroxy-3′-allylphenyl)-4,4-dimethyl-2-oxazoline
(H-Allox, 3)
Neat allyl ether 2 (3.3 g, 0.014 mol) was placed in a
Kontes flask and heated with rapid stirring at 160 °C for
36 h. The resulting brown oil was chromatographed on silica
gel (hexane – diethyl ether, 30:1) to afford 3 as a straw-
colored oil. Carrying out the Claissen rearrangement at higher
temperatures decreases the time required, but results in con-
siderable production of the 4′-allyl isomer that is difficult to
1
white crystals. Yield: 12.5 g (78%); mp 58–60 °C. H NMR
(d-chloroform) δ: 12.17 (br s, 1H, OH), 7.65 (d, 1H, 6′-
arylH, J = 7.5 Hz), 7.34 (t, 1H, 4′-arylH, J = 7.5 Hz), 6.98
(d, 1H, 3′-arylH, J = 8.0 Hz), 6.83 (t, 1H, 5′-arylH, J =
7.8 Hz), 4.07 (s, 2H, CH2), 1.37 (s, 6H, CH3). 13C NMR (d-
chloroform) δ: 164.2 (C=N), 159.9 (2′-arylC), 133.2, 127.9,
118.6, 116.7 (arylCH), 112.4 (1′-arylC), 78.4 (CMe2CH2O),
67.1 (CMe2), 28.5 (CH3).
1
remove chromatographically. Yield: 2.9 g (88%). H NMR
(d-chloroform) : 12.42 (br s, 1H, OH), 7.43 (d, 1H, 6 -arylH,
δ
′
J = 7.7 Hz), 7.18 (d, 1H, 4′-arylH, J = 7.8 Hz), 6.72 (t, 1H,
5 -arylH, J = 7.7 Hz), 5.95 (m, 1H, CH=CH ), 5.05 (d, 1H,
′
2
CH=CHtrans, J = 16.5 Hz), 5.02 (d, 1H, CH=CHcis, J =
10.1 Hz), 4.01 (s, 2H, CH2), 3.39 (d, 2H, CH2CH=CH2, J =
6.7 Hz), 1.30 (s, 6H, CH3). 13C NMR (d-chloroform) δ:
164.1 (C=N), 158.0 (2′-arylC), 136.9 (CH=CH2), 133.6
(arylCH), 127.8 (3′-arylC), 126.2, 118.4 (arylCH), 115.8
(CH=CH2), 112.4 (1′-arylC), 78.6 (CMe2CH2O), 67.3
(CMe2), 34.3 (CH2CH=CH2), 28.8 (CH3). HR-MS (EI)
calcd.: M+ 231.1259; found: 231.1265.
2-(2′-Allyloxyphenyl)-4,4-dimethyl-2-oxazoline (2)
The crude 2-(2′-hydroxyphenyl)-4,4-dimethyloxazoline
(4.0 g, 0.021 mol) was placed in a round bottomed flask and
dissolved in 100 mL of acetone. Excess allylbromide
(3.19 g, 0.0264 mol), K2CO3 (12.1 g, 0.0872 mol), and n-
Bu4N+Br– (0.34 g, 0.0017 mol), as a phase-transfer catalyst,
were added. The suspension was stirred vigorously at room
temperature for 6 days. The salts were removed by filtration
and the solvent was stripped under vacuum. The residue was
redissolved in diethyl ether (100 mL), filtered to remove fur-
ther solids, and again taken to dryness under vacuum. The
residue was columned on silica gel (hexane) and the pure
allyl ether was isolated as a yellow oil. Yield: 4.11 g (85%).
1H NMR (d-chloroform) δ: 7.61 (d, 1H, 6′-arylH, J = 7.8 Hz),
7.24 (t, 1H, 4′-arylH, J = 7.6 Hz), 6.87 (d, 1H, 3′-arylH, J =
7.9 Hz), 6.81 (t, 1H, 5′-arylH, J = 7.8 Hz), 5.95 (m, 1H,
CH=CH2), 5.46 (d, 1H, CH=CHtrans, J = 16.9 Hz), 5.17 (d,
La(Allox)3 (4)
A solution of 3 (0.069 g, 0.30 mmol) in 4 mL of toluene
was added to a solution of La[N(SiMe3)2]3 (0.062 g,
0.10 mmol) in 4 mL of toluene in the glovebox. The solution
immediately turned pale yellow and all solids dissolved. The
solution was stirred overnight at room temperature and the
solvent was removed under reduced pressure to yield a gran-
ular white solid. The solid was redissolved in a minimum of
hot hexane–toluene (80:20), filtered though a Celite® pad,
and the filtrate was then allowed to slowly cool to room tem-
© 2005 NRC Canada

452
Can. J. Chem. Vol. 83, 2005
perature. The clear, colorless blocks of 4 that deposited from
4400 Hz, 18H, CH3), 37.8 (υ1/2 = 480 Hz, 6H, CH2), –1.51
(υ1/2 = 68 Hz, 3H, alkeneH), –7.0 (υ1/2 = 180 Hz, 3H,
alkeneH), –12.1 (υ1/2 = 50 Hz, 3H, arylH), –15.3 (υ1/2 =
solution were collected by vacuum filtration and dried under
1
vacuum. Yield: 0.061 g (73%); mp 180–182 °C. H NMR
(d8-toluene) δ: 7.90 (dd, 1H, 6′-arylH, J = 7.7, 1.8 Hz), 7.20
(dd, 1H, 4′-arylH, J = 7.2, 1.8 Hz), 6.59 (t, 1H, 5′-arylH,
J = 7.5 Hz), 5.98 (m, 1H, CH=CH2), 5.00 (d, 1H,
CH=CHtrans, J = 17.1 Hz), 4.90 (d, 1H, CH=CHcis, J =
10.3 Hz), 3.47 (s, 2H, OCH2), 3.45 (d, 2H, CH2CH=CH2,
J = 6.7 Hz), 1.27 (s, 6H, CMe2). 13C NMR (d8-toluene) δ:
168.0 (C=N), 157.8 (2′-arylC), 137.6 (CH=CH2), 135.1
(arylCH), 130.9 (arylC), 115.3 (arylCH), 115.0 (CH=CH2),
112.0 (1′-arylC), 78.0 (CMe2CH2O), 67.4 (CMe2), 35.5
(CH2CH=CH2), 28.5 (CH3). The remaining aryl C resonance
was not discernible. MS (EI) m/z (amu): 829 (M+ 139La,
44%), 599 (M+ – L, 10%), 527 (M+ – L – (C4H8O), 7%),
455 (M+ – L – 2(C4H8O), 100%). Anal. calcd. for
C42H48N3O6La (%): C 60.80, H 5.83, N 5.06; found: C
61.22, H 5.52, N 5.00.
35 Hz, 3H, arylH), –20.8 (υ1/2 = 80 Hz, 3H, arylH). One
CH2 (6H) resonance was not observable. MS (EI) m/z (amu):
858 (M+ 168Er, 100%), 628 (M+ – L, 7%), 554 (M+ – L –
(C4H8O), 12%), 482 (M+ – L – 2(C4H8O), 43%). Anal.
calcd. for C42H48N3O6Er (%): C 58.79, H 5.64, N 4.90;
found: C 60.71, H 5.92, N 4.62.
Y(Allox)3 (8)
Complex 8 was isolated as colorless crystals using the
procedure described for 4 starting from Y[N(SiMe3)2]3
(0.057 g, 0.10 mmol) and 3 equiv. of 3. Yield: 0.036 g
1
(63%); mp 185 to 186 °C. H NMR (d8-toluene) δ: 7.90 (dd,
3H, 6′-arylH, J = 8.0, 1.9 Hz), 7.16 (dd, 3H, 4′-arylH, J =
7.3, 1.9 Hz), 6.57 (t, 3H, 5′-arylH, J = 7.7 Hz), 5.89 (m, 3H,
CH=CH2), 4.95 (d, 3H, CH=CHtrans, J = 17.0 Hz), 4.86 (d,
3H, CH=CHcis, J = 10.0 Hz), 3.51 (s, 6H, CH2), 3.40 (d, 6H,
CH2CH=CH2, J = 6.7 Hz), 1.25 (s, 18H, CMe2). 13C NMR
(d8-toluene) δ: 138.1 (CH=CH2), 134.8 (arylCH), 132.2
(arylC), 115.1 (arylCH), 114.7 (CH=CH2), 112.5 (1′-arylC),
78.1 (CMe2CH2O), 67.8 (CMe2), 35.4 (CH2CH=CH2), 28.4
(CH3). The remaining aryl C resonances and the oxazoline
C=N resonance were not discernible. MS (EI) m/z (amu):
779 (M+ 89Y, 100%), 405 (M+ – L – 2(C4H8O), 26%). Anal.
calcd. for C42H48N3O6Y (%): C 64.69, H 6.20, N 5.39;
found: C 65.01, H 6.44, N 5.15.
Ce(Allox)3 (5)
Complex 5 was prepared using a procedure analogous to 4
from Ce[N(SiMe3)2]3 (0.062 g, 0.10 mmol) and 3 equiv. of
3. Small, bright yellow blocks of 5 were isolated by allow-
ing a warm (60 °C) hexane–toluene solution to cool to room
temperature. Yield: 0.072 g (87%); mp 196–199 °C. 1H
NMR (d8-toluene, 323 K) δ: 11.86 (d, 3H, 6′-arylH, J =
7.4 Hz), 10.63 (d, 3H, 4′-arylH, J = 6.8 Hz), 9.37 (t, 3H, 5′-
arylH, J = 6.3 Hz), 8.06 (m, 3H, CH=CH2), 5.65 (d, 3H,
CH=CHtrans, J = 14 Hz), 5.39 (d, 3H, CH=CHcis, J = 8 Hz),
4.27 (br s, υ1/2 = 18 Hz, 6H, CH2CH=CH2), 1.00 (s, υ1/2
=
8 Hz, 6H, CH2), –9.48 (br s, υ1/2 = 34 Hz, 18H, CH3). MS
(EI) m/z (amu): 830 (M+ 140Ce, 43%), 516 (100%). Anal.
calcd. for C42H48N3O6Ce (%): C 60.71, H 5.82, N 5.06;
found: C 61.05, H 5.71, N 4.92.
X-ray crystallographic studies
Crystals of 5 and 7 suitable for X-ray crystallography
were grown by allowing a hot, concentrated, toluene–hexane
solution to slowly cool to room temperature. Typical crystal-
lographic procedures are illustrated for compound 5 in the
following section. Crystals of 5 suitable for X-ray diffraction
were removed from the flask, covered with a layer of hydro-
carbon oil, attached to a glass fiber, and placed in the low-
temperature nitrogen stream (10). Data was collected at
83(2) K using a Bruker/Siemens SMART APEX instrument
(Mo Kα radiation, λ = 0.710 73 Å) equipped with a
Cryocool NeverIce low-temperature device. Data were mea-
sured using ω scans of 0.3° per frame for 30 s, and a full
sphere of data was collected. A total of 2132 frames were
collected with a final resolution of 0.77 Å. The first 50
frames were recollected at the end of data collection to mon-
itor for decay. Cell parameters were retrieved using SMART
(11) software and refined using SAINTPlus (12) on all ob-
served reflections. Data reduction and correction for
Lorentzian polarization and decay were performed using the
SAINTPlus (12) software. Absorption corrections were ap-
plied using SADABS (13). The structure was solved by di-
rect methods and refined by the least-squares method on F2
using the SHELXTL program package (14). The structure
was solved in the space group P21/n (No. 14) by analysis of
systematic absences. All atoms were refined anisotropically.
One of the terminal vinyl groups was disordered. This was
modeled with an occupancy of 66% for the major fraction.
Soft restraints were applied to maintain the correct geome-
try. No decomposition was observed during data collection.
Details of the data collection and refinement are given in Ta-
Sm(Allox)3 (6)
Complex 6 was isolated as colorless crystals in the same
manner as 4 and 5 previously starting from Sm[N(SiMe3)2]3
(0.063 g, 0.10 mmol) and 3 equiv. of 3. Yield: 0.052 g
(62%); mp 173–176 °C. IR (Nujol, NaCl, cm–1): 1652 (vw),
1635 (sh w), 1597 (s), 1558 (w), 1345 (w), 1260 (s), 1226
(vw), 1190 (w), 1160 (w), 1132 (m), 1096 (m), 1065 (w),
1055 (w), 862 (w), 822 (w), 768 (m), 750 (w), 720 (w), 667
(m). The IR spectra of complexes 4, 5, 7, and 8 were
1
superimposable with that of 6. H NMR (d8-toluene, 373 K)
δ: 9.23 (d, 3H, 6′-arylH, J = 8.7 Hz), 7.87 (d, 3H, 4′-arylH,
J = 6.5 Hz), 7.68 (t, 3H, 5′-arylH, J = 8.5 Hz), 6.10 (m, 3H,
CH=CH2), 4.97 (d, 3H, CH=CHtrans, J = 16 Hz), 4.85 (d,
3H, CH=CHcis, J = 11 Hz), 2.83 (d, 6H, CH2CH=CH2, J =
6.4 Hz), 2.73 (s, 6H, CH2), –2.94 (br s, υ1/2 = 4 Hz, 18H,
CH3). MS (EI) m/z (amu): 842 (M+ 152Sm, 68%), 612 (M+ –
L, 10%), 540 (M+ – L – (C4H8O), 16%), 468 (M+ – L –
2(C4H8O), 100%). Anal. calcd. for C42H48N3O6Sm (%): C
59.97, H 5.75, N 5.00; found: C 60.30, H 5.58, N 5.01.
Er(Allox)3 (7)
Complex 7 was obtained as well-formed, bright pink
prisms using the procedure outlined for 4 previously starting
from Er[N(SiMe3)2]3 (0.065 g, 0.10 mmol) and 3 equiv. of 3.
Yield: 0.077 g (90%); mp 168–170 °C. 1H NMR (d8-toluene,
373 K, all resonances are broad singlets) δ: 80.1 (υ1/2
=
© 2005 NRC Canada

Berg et al.
453
Table 1. Crystallographic data for Ln(Allox)3 (5 (Ce), 7 (Er)).
Empirical formula
C42H48CeN3O6
C42H48ErN3O6
Formula weight
Temperature (K)
Wavelength (Å)
Crystal system
Space group
831.0
83
0.710 73
Monoclinic
P21/n (No. 14)
858.1
223
0.710 73
Monoclinic
P21/n (No. 14)
a (Å)
b (Å)
c (Å)
14.641 1(10)
15.057 6(10)
17.102 3(11)
94.976(1)
3756.2(4)
4
1.47
1.264
1708
14.426(3)
15.186(3)
17.065(4)
93.849(4)
3781.5(13)
4
1.51
2.27
1748
β (°)
Volume (Å3)
Z
Density (calcd., g/cm3)
Absorption coeff. (µ, mm–1)
F(000)
Crystal size (mm)
Crystal color and habit
Diffractometer
0.13 × 0.08 × 0.07
Yellow needle
Bruker Smart APEX
1.80–27.50
48 791, 8 624
0.92 and 0.85
8 624 / 7 / 482
1.01
0.16 × 0.17 × 0.37
Pink needle
Bruker Smart 1000
2.0–25.0
19 548, 6 679
1.00 and 0.85
6 679 / 2 / 488
0.92
θ range (°)
Reflections collect., independent
Max and min transmission
Data / restraints / parameters
Goodness-of-fit on F2
Final R indices (I > 2σ(I))
R (all data)
R1 = 0.045, wR2 = 0.093
R1 = 0.092
2.351 and –0.745
R1 = 0.024, wR2 = 0.046
R1 = 0.057
0.50 and –0.29
Largest diff. peak and hole (e/Å3)
ble 1. Further details are provided in the Supplementary ma-
terial.3
carbon percentage was slightly high in all cases, possibly in-
dicating retention of toluene solvent (from recrystallization).
All of the Ln(Allox)3 complexes prepared gave well-
behaved mass spectra with molecular ions as the base peaks
for complexes of the smaller metals (Er 7 and Y 8). Simula-
tion of the molecular ion pattern provided unequivocal evi-
dence for the formation of the tris chelate complexes,
especially for 4–7, since these metals have distinctive isoto-
pic patterns. An example of the match between the experi-
mental and calculated molecular ion pattern is shown in
Fig. 1 for Sm(Allox)3 (6). The dominant fragments in the
mass spectra of 4–8 represent the loss of one or two ligands
and the loss of C4H8O, presumably from cleavage of the
oxazoline ring. The latter has been observed previously dur-
ing the thermal decomposition of Ce(Me2-phenox)4 (Me2-
phenox, ligand I) (5).
Results and discussion
Synthesis and characterization
The chelating neutral ligand 2-(2′-hydroxy-3′-allylphenyl)-
4,4-dimethyl-2-oxazoline (HAllox, 3) was prepared in three
steps from 2-cyanophenol as illustrated in Scheme 1. The
oxazoline ring was first introduced by the Lewis-acid-
catalyzed cyclization of 2-cyanophenol and 2-amino-2-
methylpropanol using ytterbium triflate as the catalyst. Al-
though this reaction proceeds slowly at 170 °C in 1,2-
dichlorobenzene, we prefer this route to the ZnCl2-catalyzed
cyclization (9) because the product is cleaner and yields are
comparable. Introduction of the 3′-allyl group was achieved
in two steps by the Claissen rearrangement of the allyl
phenyl ether 2.
Solid-state structures
Crystallographic data for 5 and 7 are summarized in Ta-
ble 1. Selected distances and angles are given in Table 2, and
ORTEP3 (15) plots of the structures are shown in Figs. 2 (5,
Ce) and 3 (7, Er). The structures of 5 and 7 are very similar
to one another, showing a distorted octahedral coordination
environment and a mer arrangement of the three chelating
Allox ligands. The allyl groups clearly are not within bond-
ing distance of the metal as the closest contacts between the
internal or terminal alkene carbons and the metals are more
than 4 Å.
More than 40 structures have been reported of metal com-
plexes containing derivatives of the 2-oxazolylphenoxide
chelate unit. These structures span all groups of the d-block
and the Group 13 metals, but there are no examples contain-
Lanthanide complexes 4–8 of the new oxazoline-phenoxide
ligand were most readily prepared by an acid–base reaction
between the neutral ligand 3 and Ln[N(SiMe3)2]3 in toluene
(Scheme 1: Ln = La 4, Ce 5, Sm 6, Er 7, Y 8). Metathesis
reactions between the lithium salt LiAllox and YCl3 or
LaCl3 in THF produced 4 or 8, respectively, but the yields
were much lower by this route (20%–30%). The complexes
displayed the characteristic colour of each ion (Ce yellow,
Er pink, La, Sm, Y colourless). Complexes 4–8 display simi-
lar melting points and their IR spectra as Nujol mulls are
superimposable suggesting that these complexes are
isostructural in the solid state. This observation was con-
firmed for 5 and 7 by X-ray crystallography (vide infra). All
complexes gave an acceptable elemental analysis, but the
© 2005 NRC Canada

454
Can. J. Chem. Vol. 83, 2005
Scheme 1.
°C
170 , 24 h
Cl
OH
N
O
N
O
OH
Br
CN
O
Cl
1 mol%
NBun + Br-, K2CO3
4
1
2
Yb(CF3SO3)3
acetone,
ambient, 6 days
OH
H2N
160 °C
36 h
OH
N
O
N
Ln[N(SiMe3)2]3
toluene
O
4 (La), 5 (Ce), 6 (Sm)
7 (Er), 8 (Y)
Ln
H-Allox
3,
3
O
Table 2. Selected distances and angles for Ln(Allox)3 (5 (Ce), 7 (Er)).a
5
7
Bond distances (Å)
Ce(1)—N(1A)
Ce(1)—N(1B)
Ce(1)—N(1C)
Ce(1)—O(1A)
Ce(1)—O(1B)
Ce(1)—O(1C)
2.697(3)
2.584(4)
2.577(4)
2.260(3)
2.251(3)
2.275(3)
Er(1)—N(1)
Er(1)—N(2)
Er(1)—N(3)
Er(1)—O(1)
Er(1)—O(3)
Er(1)—O(5)
2.420(3)
2.433(3)
2.541(3)
2.148(2)
2.124(3)
2.140(2)
Bond angles (°)
Ce(1)-O(1A)-C(1A)
Ce(1)-O(1B)-C(1B)
Ce(1)-O(1C)-C(1C)
N(1A)-Ce(1)-N(1B)
N(1A)-Ce(1)-N(1C)
N(1A)-Ce(1)-O(1A)
N(1A)-Ce(1)-O(1B)
N(1A)-Ce(1)-O(1C)
N(1B)-Ce(1)-N(1C)
N(1B)-Ce(1)-O(1A)
N(1B)-Ce(1)-O(1B)
N(1B)-Ce(1)-O(1C)
N(1C)-Ce(1)-O(1A)
N(1C)-Ce(1)-O(1B)
N(1C)-Ce(1)-O(1C)
O(1A)-Ce(1)-O(1B)
O(1A)-Ce(1)-O(1C)
O(1B)-Ce(1)-O(1C)
148.9(3)
147.3(3)
146.2(3)
Er(1)-O(1)-C(1)
Er(1)-O(3)-C(15)
Er(1)-O(5)-C(29)
N(1)-Er(1)-N(2)
N(1)-Er(1)-N(3)
N(1)-Er(1)-O(1)
N(1)-Er(1)-O(3)
N(1)-Er(1)-O(5)
N(2)-Er(1)-N(3)
N(2)-Er(1)-O(1)
N(2)-Er(1)-O(3)
N(2)-Er(1)-O(5)
N(3)-Er(1)-O(1)
N(3)-Er(1)-O(3)
N(3)-Er(1)-O(5)
O(1)-Er(1)-O(3)
O(1)-Er(1)-O(5)
O(3)-Er(1)-O(5)
142.7(2)
143.9(2)
146.4(2)
160.7(9)
91.02(10)
75.41(9)
90.72(9)
100.61(9)
105.77(10)
94.46(9)
75.14(10)
93.50(9)
90.87(9)
166.86(9)
73.24(9)
102.16(9)
163.67(9)
93.65(9)
107.79(11)
92.92(11)
68.62(11)
161.41(11)
92.23(11)
156.48(12)
95.78(11)
69.94(11)
97.49(10)
102.22(11)
93.90(11)
70.02(11)
93.01(11)
159.36(11)
106.36(11)
aEstimated standard deviations are in parentheses.
ing Group 3, lanthanide, (or actinide) ions.4 Of the published
structures, the best comparisons to 5 and 7 are the six-
coordinate, tris chelate complexes of Al, Ga, and In, all of
which also adopt a mer geometry (16). In fact, the other
known tris chelates of this ligand type, three for Mn (2d, 2h)
and one for Fe (17), crystallize exclusively in the mer
geometry. There appears to be a pronounced trans influence
operative in 5 and 7 as the M—N distance trans to oxygen is
ca. 0.12 Å longer than that trans to nitrogen in both cases. A
similar, but much smaller, effect is observed in the tris che-
4 A complete listing of these structures and a table summarizing important metrical parameters is given in the Supplementary material.3
© 2005 NRC Canada

Berg et al.
455
Fig. 3. ORTEP3 (16) plot (30% probability) of Er(Allox)3 (7).
Fig. 2. ORTEP3 (16) plot (30% probability) of Ce(Allox)3 (5).
O2
O2C
N1C
C1C
O1C
N1
C1
O1
O6
Er1
O2A
C1A
Ce1
N3
N1A
O1A
O3
O5
C29
O1B
C1B
C15
N2
N1B
O4
O2B
tively short compared to that in 5. Since Er is the smaller of
the two metals, this runs counter to steric arguments; stron-
ger oxazoline binding owing to the higher charge to size ra-
tio of Er may explain this effect. In any case, it seems quite
clear from the foregoing discussion that these complexes are
not significantly crowded, as we might expect for a six-
coordinate lanthanide center and relatively small ligand sub-
stituents.
The Ln-O-C angles of the phenoxide group are consider-
ably less bent in 5 (146.2(3)°–148.9(3)°) and 7 (142.7(2)°–
146.4(2)°) than those in the Group 13 tris chelates (122°–
135°) (16), and only complexes of the Group 4 metals show
M-O-C angles of more than 140° (19). Terminal Ln-O-
C(aryl) angles are usually greater than 160° (20) so the ob-
served angles are in fact more bent than usual because of the
constraints of the chelate geometry.6 The chelate bite itself
(O-Ln-N (5 (68.62(11)°–70.02(11)°), 7 (73.24(9)°–75.41(9)°))
is the smallest of any reported 2-oxazolylphenoxide struc-
ture (range (76°–96°), median (87°)),4 but this is primarily a
reflection of the fact that Ce and Er are the largest metals to
which this ligand has been complexed.
late complexes of Group 13 where the average difference is
only ca. 0.03 Å (16). The Ln—O bond lengths in 5 and 7 are
0.07 Å shorter than predicted from In(Phox)3 (PhoxH = 2-
(2′-hydroxyphenyl)-2-oxazoline) (16a) after adjustment for
differences in ionic radii5 (18). In contrast, the Ln—N dis-
tances are 0.20 (5, Ce) and 0.17 (7, Er) Å longer than ex-
pected using this same comparison. All of these observations
are consistent with a greater affinity for oxygen vs. nitrogen
for the lanthanides.
There are only two structures reported of any type that
contain a neutral oxazoline ring bonded to a Group 3 or
lanthanide ion and both complexes contain the oxazoline
ring as part of a chiral, tridentate “pybox” ligand (Sc(Ph-
pybox)(OTf)3(H2O) (4c) and [La(Ph2-pybox)(OTf)2(H2O)4]+-
[OTf]– (4b)). The Sc—N(ox) distances in the seven-
coordinate Sc complex predict Ce—N and Er—N distances
of 2.56 and 2.44 Å, respectively, almost exactly matching
those observed in 5 and 7 for the trans pair of oxazoline
Ln—N bonds.5 The La—N(ox) distances in the nine-
coordinate La complex predict Ce—N and Er—N distances
of 2.47 and 2.35 Å, respectively, far shorter than those actu-
ally observed. However, the La comparison is based on the
nine-coordinate radius of La3+ and since the La centre is
bound by four small water molecules, this radius is likely an
overestimation of the size of the metal ion.
Solution behaviour
The NMR spectra of 4–8 in d8-toluene each show a single
set of nine resonances for the Allox ligand consistent with
average C3v molecular symmetry. Since this symmetry is un-
likely in a static structure, we investigated the variable-
Subtle differences are detectable between the structures of
5 and 7. In particular, although the difference in Ln—O
bond lengths in the two structures (0.125 Å) matches the dif-
ference in ionic radii for the two metals (18), the difference
in Ln—N bond distance (0.155 Å) is slightly larger than ex-
pected. Thus, the Er—N(oxazoline) distance in 7 is rela-
1
temperature H NMR spectra of all complexes at both high
and low temperature. In all cases, lowering the temperature
resulted in decoalescence into at least 17, and usually more,
resonances (vide infra). The resonances for paramagnetic
complexes 5–7 showed good linear behaviour in δ vs. 1/T
5 Ionic radii from ref. 18: coordination number (CN) = 6 for In3+ 0.800 Å, Ce3+ 1.01 Å, Y3+ 0.90 Å, Er3+ 0.89 Å; CN = 7, Sc3+ 0.81 Å; CN =
9, La3+ 1.22 Å.
6 The mean Ln-O-C angle in terminal phenoxides bearing only H or Me groups in the ortho positions is 167° with a standard deviation of 8°
and an overall range of 144°–180°. Some typical examples are given in ref. 20.
© 2005 NRC Canada

456
Can. J. Chem. Vol. 83, 2005
1
1
Fig. 4. H NMR chemical shift (δ) vs. the inverse temperature
(1/T) plot for Er(Allox)3 (7) above coalescence (d8-toluene,
360 MHz).
Fig. 5. A portion of the low temperature H NMR spectrum of
Sm(Allox)3 (6, d8-toluene, 500 MHz, diamagnetic peaks between
0 and 1 ppm suppressed).
90
A
80
A
E
70
H
O
H
N
296 K
206 K
186 K
C
60
50
40
30
20
10
0
B
H
F
O
B
H
D
G
I
E
I
-10
-20
-30
-40
F
C
G
H
0.0026 0.0028 0.0030 0.0032 0.0034 0.0036
1/T (K -1
-2.0
-6.0 -10.0 -14.0 -18.0
ppm
)
plots above and below coalescence, consistent with a single
species in each temperature regime. A sample plot for 7
above coalescence is shown in Fig. 4; remaining plots are
included in the Supplementary material.3 The line widths be-
low coalescence were quite broad for Er, but the wider
chemical shift window made it possible to distinguish 30
peaks in the low-temperature spectra. This is consistent with
the low symmetry environment found in the solid-state
structure since 36 resonances are expected if the molecule
possesses C1 symmetry.
1
Fig. 6. Variable-temperature H NMR spectra of Y(Allox)3 (8)
(d8-toluene, 360 MHz, T = toluene, * = impurity).
T
In the case of Ce and Sm, resonances clearly overlap with
one another, but the observation of six methyl signals in the
low-temperature spectra of 6 indicates that the low-
temperature limiting structure of this complex also possesses
C1 symmetry (Fig. 5). Addition of d8-THF to a d8-toluene
solution of 6 had a minor effect on the chemical shifts ob-
served, but did not alter the coalescence behaviour.
Changing the concentration of 6 over a fivefold range also
caused no changes in dynamic behaviour. Taken together,
the observation of linear δ vs. 1/T behaviour above or below
coalescence and the lack of concentration or coordinating
solvent effects strongly imply that an intramolecular flux-
ional process is operative.
T
293 K
*
253 K
233 K
223 K
The diamagnetic complexes 4 and 8 show similar behav-
iour to their paramagnetic counterparts except that the smaller
spectral dispersion leads to a much lower coalescence tem-
perature and more overlap in the low-temperature spectra.
The full VT NMR spectra of 8 are shown in Fig. 6 and an
expansion of the most downfield aryl resonance is shown in
Fig. 7. Most of the resonances that have decoalesced in the
193 K spectrum (Fig. 6) appear to be in 2:1 intensity ratio
(resonances near 6.0 and 8.2 ppm, for example). However,
closer inspection reveals that this is most likely due to acci-
dental overlap of resonances. This is clear in the expansion
of the downfield aryl resonance (6′-aryl H) because this pro-
ton cannot give rise to a true triplet given the coupling con-
stants involved for this aryl position. The observation of an
213 K
193 K
8.0 7.2 6.4 5.6 4.8 4.0 3.2 2.4 1.6 0.8
ppm
© 2005 NRC Canada

Berg et al.
457
Fig. 7. Expansion of the downfield aryl region of Fig. 6.
Fig. 8. Experimental and simulated (DNMR3) lineshapes for the
6′-aryl resonance of Y(Allox)3 (8) at selected temperatures.
223 K
k = 110 s-1
296 K
213 K
193 K
k = 40 s-1
k = 9.5 s-1
253 K
233 K
Fig. 9. Eyring plot for the exchange process of Y(Allox)3 (8).
2
1
0
223 K
213 K
203 K
-1
Slope = -2850
Intercept = 11.8
-2
-3
0.0035 0.0040 0.0045 0.0050
1/T (K-1
193 K
)
8.60
8.40
8.20
ppm
8.00
7.80
ment of the anion in K3[Ga(catecholate)3] has been estab-
lished that shows a negative activation entropy of –20 to
–47 J mol–1 K–1 depending on the solvent and conditions (23).
The foregoing discussion suggests that 4–8 undergo a six-
coordinate fluxional process in solution. The options include
the familiar Bailar (trigonal) or Rây–Dutt (rhombic) twists
(Figs. 10 and 11) (24). Negative entropies of activation are
often associated with both of these processes (23, 25).
The Bailar twist cannot interconvert mer and fac isomers,
but it can scramble ligands between sites within the mer ge-
ometry and cause racemization (mer-∆ with mer-Λ). The
Rây–Dutt twist can interconvert all four isomers (mer-∆,
mer-Λ, fac-∆, fac-Λ) (26). We cannot distinguish between
these two possibilities because we would need to observe fac
and mer exchange to confirm a Rây–Dutt twist (26) and the
fac isomer is not observed in the low-temperature NMR
spectra of 4–8.
Theoretically, it is also possible to distinguish between
these two mechanisms kinetically since the Bailar twist ex-
changes ligand environments at different rates [A → C, B →
B, C → A] while a Rây–Dutt twist permutes ligand environ-
ments at an equal rate [A → C, B → A, C → B] (Figs. 10a–
10c) (26, 27). Unfortunately, the spectral dispersion in the
VT NMR of diamagnetic complexes 4 and 8 is not sufficient
to model exchange by anything more than a single process.
apparent triplet can only be explained by the overlap of two
doublets (a doublet of doublets pattern is expected, but the
smaller coupling is not resolvable given the line widths in
this spectrum). The four methyl resonances between 1.0–
1.8 ppm in the 193 K spectrum integrate as 3:3:3:9, which
also can only be explained by accidental overlap of three (of
the six) inequivalent methyl groups at 1.0 ppm.
The VT behaviour of the downfield aryl resonance was
simulated by DNMR3 assuming that a single process ex-
changes all three inequivalent resonances; a subset of the
simulated and observed spectra is shown in Fig. 8. The
Eyring plot for this process (Fig. 9) gave ∆H‡ = 24
2 kJ mol–1 and ∆S‡ = –99 15 J mol–1 K–1. While the errors
in these parameters are admittedly quite large, the large neg-
ative entropy of activation is inconsistent with a dissociative
mechanism (21). An associative process to produce a seven-
coordinate complex is also highly unlikely given the lack of
solvent and concentration dependence. A process involving a
“hinged” five-coordinate intermediate cannot be entirely ex-
cluded, but most examples of rearrangements proceeding by
partial dissociation of a chelate ligand show positive activa-
tion entropies (22). Conversely, a six-coordinate rearrange-
© 2005 NRC Canada

458
Can. J. Chem. Vol. 83, 2005
Fig. 10. Interconversion of geometric and optical isomers by the
Rây–Dutt (rhombic) twist mechanism.
on mer-Ga(fox)3 (Hfox = 5-fluoro-8-hydroxyquinoline)
using 2D EXSY NMR was able to kinetically distinguish be-
tween the two twist mechanisms and found that both con-
tributed to the fluxional behaviour of this compound (27).
Interestingly, the b/l ratio for mer-Ga(ox)3 (Hox = 8-
hydroxyquinoline) is 0.90 (29), almost exactly the value pre-
dicted by Rodger and Johnson to give equal energetics for
either twist mechanism. Thus, while theoretical treatments
suggest that 4–8 should undergo a Bailar rather than a Rây–
Dutt twist, the mechanisms may be close enough in energy
that both are operative.8
The fluxional Group 13 tris chelates (M(3′-MePhox)3,
M = Al, Ga, In) were previously investigated by VT NMR
spectroscopy (16a). These complexes also undergo intra-
molecular rearrangement of inequivalent ligands within the
mer geometry on the NMR timescale at, or just above, room
temperature. In the case of In, there was some evidence for
formation of the fac isomer as a distinct (nonexchanging)
species at high temperature. The authors did not determine
∆S‡ and were unable to rule out a bond rupture (five-
coordinate Berry pseudorotation) mechanism. However, ∆G‡
values were calculated at the coalescence temperatures and
as might be expected, they decrease in the order Al
(72 kJ mol–1) > Ga (64 kJ mol–1) > In (<42 kJ mol–1). The
value reported for In was based on the lowest temperature
attained of 213 K; using the activation parameters for 8, we
calculate a value of ∆G‡ = 45 kJ mol–1 at this temperature.
While it might seem surprising that the much larger4 Y com-
plex has a larger ∆G‡ value, the oxazoline rings in the In
structure are unsubstituted while those in 8 bear two methyl
groups in the 4 position. These methyl groups are relatively
close to the metal centre (avg. 3.89 Å in 7) and may well in-
crease the barrier to a six-coordinate rearrangement.
Concluding remarks
Fig. 11. The trigonal prismatic Rây–Dutt intermediate.
Lanthanide complexes of the hybrid oxazoline-phenoxide
chelating ligand Allox are readily accessible and adopt a six-
coordinate mer geometry in the solid state without coordina-
tion of the pendant allyl substituents. NMR evidence sug-
gests that this structure is maintained in solution over a wide
temperature range, but exchange of inequivalent ligands
within the mer geometry occurs rapidly on the NMR
timescale at room temperature for all complexes studied. A
more detailed study of the exchange process for Y complex
8 supports an intramolecular six-coordinate rearrangement,
although it is not possible to distinguish between a Bailar
(trigonal) or Rây–Dutt (rhombic) twist mechanism.
N
O
N
N
O
O
If the fac isomer is present and exchanging with the mer iso-
mer, it must be in low concentration at all temperatures or
we would not observe linear behaviour in plots of δ vs. 1/T
for 5–7.7
The relative energetics of the Rây–Dutt and Bailar twist
mechanisms have been assessed theoretically by Rodger and
Johnson (28). They suggest that the Bailar twist will gener-
ally be favoured if the ratio of chelate bite angle (b) to hard
sphere contact distance (l) between adjacent ligand atoms
(not part of the same chelate) is small (b/l ≈ 0.5), and the
Rây–Dutt twist will be favoured if this ratio is large (b/l ≈
1.5). According to these calculations, there is no energetic
difference between the two mechanisms at b/l = 0.914. This
ratio is 0.76 (Ce) and 0.82 (Er) in complexes 5 and 7. It
should also be pointed out that a recent experimental study
Acknowledgment
The support of the Natural Sciences and Engineering Re-
search Council of Canada (NSERC) is gratefully acknowl-
edged.
References
1. (a) A. Chimiak and J.B. Nielands. Struct. Bonding (Berlin),
58, 90 (1984); (b) R.J. Bergeron. Chem. Rev. 84, 587 (1984);
7 Assuming that the ratio of fac and mer isomers varies significantly with temperature.
8 For a discussion of electronic factors in six-coordinate rearrangements see ref. 30.
© 2005 NRC Canada

Berg et al.
(c) P.J. Maurer and M.J. Miller. J. Am. Chem. Soc. 105, 240
459
15. L.J. Farrugia. J. Appl. Crystallogr. 30, 565 (1997).
(1983); (d) Y. Nagao, T. Miyasaka, T. Hagiwara, and E.
Fujita. J. Chem. Soc. Perkin Trans. 1, 183 (1984).
16. (a) H.R. Hoveyda, V. Karunaratne, S.J. Rettig, and C. Orvig.
Inorg. Chem. 31, 5408 (1992); (b) H.R. Hoveyda, C. Orvig,
and S.J. Rettig. Acta Crystallogr. Sect. C, C50, 1906 (1994).
17. H. Kooijman, A.L. Spek, M. Hoogenraad, E. Bouwman, J.G.
Haasnoot, and J. Reedjik. Acta Crystallogr. Sect. C, C58, 390
(2002).
18. R.D. Shannon. Acta Crystallogr. Sect. A, A32, 751 (1976).
19. (a) P.G. Cozzi, C. Floriani, A. Chiesi-Villa, and C. Rizzoli.
Inorg. Chem. 34, 2921 (1995); (b) P.G. Cozzi, E. Gallo, C.
Floriani, A. Chiesi-Villa, and C. Rizzoli. Organometallics, 14,
4994 (1995).
20. (a) W.J. Evans, T.P. Hanusa, and K.R. Levan. Inorg. Chim.
Acta, 110, 191 (1985); (b) W.J. Evans, M.A. Ansari, J.W.
Ziller, and S.I. Khan. J. Organomet. Chem. 553, 141 (1998);
(c) W.J. Evans, J.M. Olafson, and J.W. Ziller. Inorg. Chem. 28,
4308 (1989); (d) W.J. Evans, M.A. Greci, and J.W. Ziller. J.
Chem. Soc. Chem. Comm. 2367 (1998); (e) G.R. Giesbrecht,
J.C. Gordon, D.L. Clark, B.L. Scott, J.G. Watkin, and K.J.
Young. Inorg. Chem. 41, 6372 (2002); (f) S. Daniele, L.
Hubert-Pfalzgraf, and J. Vaissermann. Polyhedron, 22, 127
(2003); (g) W.J. Evans, M.A. Ansari, and J.W. Ziller. Inorg.
Chem. 34, 3079 (1995); (h) H.C. Aspinall and M. Williams.
Inorg. Chem. 35, 255 (1996); (i) M.E.G. Skinner, B.R. Tyrrell,
B.D. Ward, and P. Mountford. J. Organomet. Chem. 647, 145
(2002); (j) A.C. Hiller, S.-Y. Liu, A. Sella, and M.R.J. Elsegood.
Inorg. Chem. 39, 2635 (2000); (k) G.B. Deacon, C.M. Forsyth,
and N.M. Scott. Eur. J. Inorg. Chem. 2501 (2000).
2. (a) C. Bolm, G. Schlingloff, and K. Wieghardt. Angew. Chem.
Int. Ed. Engl. 33, 1848 (1994); (b) C. Bolm and G.
Schlingloff. J. Chem. Soc. Chem. Comm. 1247 (1995); (c) M.
Gómez-Simón, S. Jansat, G. Muller, D. Panyella, M. Font-
Bardia, and X.-J. Solans. J. Chem. Soc. Dalton Trans. 3755
(1997); (d) M. Hoogenraad, K. Ramkisoening, H. Kooijman,
A.L. Spek, E. Bouwman, J.G. Haasnoot, and J. Reedjik. Inorg.
Chim. Acta, 279, 217 (1998); (e) M. Hoogenraad, K.
Ramkisoening, H. Kooijman, A.L. Spek, E. Bouwman, J.G.
Haasnoot, and J. Reedjik. Inorg. Chim. Acta, 320, 117 (2001);
(f) G. Mugesh, H.B. Singh, and R.J. Butcher. Eur. J. Inorg.
Chem. 669 (2001); (g) M. Gómez, S. Jansat, G. Muller, G.
Noguera, H. Teruel, V. Molines, E. Cerrada, and M.
Hursthouse. Eur. J. Inorg. Chem. 1071 (2001); (h) M.
Hoogenraad, K. Ramkisoening, S. Gorter, W.L. Driessen, E.
Bouwman, J.G. Haasnoot, J. Reedjik, T. Mahabiersing, and F.
Hartl. Eur. J. Inorg. Chem. 377 (2002); (i) W. Zhao, Y. Qian, J.
Huang, and J. Duan. J. Organomet. Chem. 689, 2614 (2004);
(j) J. Arias, C.R. Newlands, and M.M. Abu-Omar. Inorg.
Chem. 40, 2185 (2001); (k) M. Gómez, G. Muller, and M.
Rocamora. Coord. Chem. Rev. 193–195, 769 (1999); (l) H.A.
McManus and P.J Guiry. Chem. Rev. 104, 4151 (2004).
3. C. Qian and L. Wang. Tetrahedron Lett. 41, 2203 (2000).
4. (a) C. Qian and L. Wang. Tetrahedron: Asymmetry, 11, 2347
(2000); (b) G. Desmoni, G. Faita, S. Filippone, M. Mella,
M.G. Zampori, and M. Zema. Tetrahedron, 57, 10 203 (2001);
(c) D.A. Evans, Z.K. Sweeney, T. Rovis, and J.S. Tedrow. J.
Am. Chem. Soc. 123, 12 095 (2001); (d) G. Desimoni, G.
Faita, and P. Quadrelli. Chem. Rev. 103, 3119 (2003); (e) G.
Desimoni, G. Faita, M. Guala, and C. Pratelli. J. Org. Chem.
68, 7862 (2003).
21. R.B. Jordan. Reaction mechanisms of inorganic and organo-
metallic systems. 2nd ed. Oxford University Press, New York.
1998. pp. 102–133.
22. (a) C.P. Casey, J.F. Klein, and M.A. Fagan. J. Am. Chem. Soc.
122, 4320 (2000); (b) T.K. Hollis, Y.J. Ahn, and F.S. Tham. J.
Chem. Soc. Chem. Comm. 2996 (2002); (c) H. Mutsuzaki, H.
Kawaguchi, P. Voth, K. Noda, S. Itoh, H.D. Takagi, K.
Kashiwabara, and K. Tatsumi. Inorg. Chem. 42, 5320 (2003).
23. B. Kersting, J.R. Telford, M. Meyer, and K.N. Raymond. J.
Am. Chem. Soc. 118, 5712 (1996).
5. H. Guillon, S. Daniele, L.G. Hubert-Pfalzgraf, and J.M. Letoffe.
Polyhedron, 23, 1467 (2004).
6. H. Kodama, J. Ito, K. Hori, T. Ohta, and I. Furakawa. J.
Organomet. Chem. 603, 6 (2000).
7. S. Hong, S. Tian, M.V. Metz, and T.J. Marks. J. Am. Chem.
24. (a) J.C. Bailar, Jr. J. Inorg. Nucl. Chem. 8, 165 (1958); (b) P.C.
Rây and N.K. Dutt. J. Indian Chem. Soc. 20, 81 (1943);
(c) M.L. Tobe and J. Burgess. Inorganic reaction mechanisms.
Adison Wesley Longman, Ltd., New York. 1999. pp. 224–239.
25. (a) L. Kakaliou, W.J. Scanlan, B. Qian, S.W. Baek, and M.R.
Smith III. Inorg. Chem. 38, 5964 (1999); (b) M. Rahim, N.J.
Taylor, S. Xin, and S. Collins. Organometallics, 17, 1315
(1998); (c) D.J. Darensbourg and R.L. Gray. Inorg. Chem. 23,
2993 (1984).
Soc. 125, 14 768 (2003).
8. D.C. Bradley, J.S. Ghotra, and F.A. Hart. J. Chem. Soc. Dalton
Trans. 1021 (1973).
9. (a) K.M. Button, R.A. Gossage, and R.K.R. Phillips. Synth.
Commun. 32, 363 (2002); (b) J.T. Banks, K.M. Button, R.A.
Gossage, T.D. Hamilton, and K.E. Kershaw. Heterocycles, 55,
2251 (2001); (c) K.M. Button and R.A. Gossage. J. Hetero-
cyclic Chem. 40, 513 (2003); (d) H. Witte and W. Seeliger.
Angew. Chem. Int. Ed. Engl. 11, 287 (1972); (e) H. Witte and
W. Seeliger. Justus Liebigs Ann. 996 (1974); (f) C. Bolm, K.
Weickhardt, M. Zehnder, and T. Ranff. Chem. Ber. 124, 1173
(1991); (g) C. Bolm, K. Weickhardt, M. Zehnder, and D.
Glasmacher. Helv. Chim. Acta, 74, 717 (1991).
10. H. Hope. Prog. Inorg. Chem. 41, 1 (1995).
11. SMART. Version 5.626 [computer program]. Bruker AXS,
Madison, Wisconsin. 2002.
12. SAINTPlus. Version 6.22 [computer program]. Bruker AXS,
Madison, Wisconsin. 2001.
26. R.C. Fay and T.S. Piper. Inorg. Chem. 3, 348 (1964).
27. M. Gromova, O. Jarjayes, S. Hamman, R. Nardin, C. Béguin,
and R. Willem. Eur. J. Inorg. Chem. 545 (2000).
28. A. Rodger and B.F.G. Johnson. Inorg. Chem. 27, 3602 (1988).
29. Calculated from the reported crystal structure data: (a) Y. Wang,
W. Zhang, Y. Li, L. Ye, and G. Yang. Chem. Mater. 11, 530
(1999); (b) H. Schmidbaur, J. Lettenbauer, D.L. Wilkinson, G.
Muller, and O. Kumberger. Z. Naturforsch. B: Chem. Sci. 46,
901 (1991); (c) L. Peck, Y. Bankovsky, V. Bel’sky, E. Silina, J.
Ashaks, and A. Sturis. Latv. J. Chem. 3, 3 (2000). CCDC
code: YINRIW01.
13. G.M. Sheldrick. SADABS. Version 2.01 [computer program].
Bruker AXS, Madison, Wisconsin. 2001.
14. G.M. Sheldrick. SHELXTL. Version 6.10 [computer program].
Bruker AXS, Madison, Wisconsin. 2001.
30. C.D. Montgomery and C.J. Shorrock. Inorg. Chim. Acta, 328,
259 (2002).
© 2005 NRC Canada