Mechanistic insight into the lanthanide (III) salt catalysed monoacylation of symmetrical diols from structural models

Article abstract of DOI:10.1039/b308171k

Model studies are presented that suggest the mechanism of the lanthanide(III) salt catalysed mono acylation of symmetrical diols proceeds via chelation of the diol and the anhydride to the lanthanide salt, followed by an 'intramolecular' acyl transfer.

Full text of DOI:10.1039/b308171k

Mechanistic insight into the lanthanide (III) salt catalysed monoacylation
of symmetrical diols from structural models†
Paul A. Clarke,*a Polly L. Arnold,*a Martin A. Smith,a Louise S. Natrajan,a Claire Wilsona and
Chuen Chanb
a School of Chemistry, University of Nottingham, University Park, Nottingham, UK NG7 2RD.
E-mail: paul.clarke@nott.ac.uk; polly.arnold@nott.ac.uk; Fax: 44 115 9513564; Tel: 44 115 9513566
b Medicinal Chemistry, GlaxoSmithKline, Gunnels Wood Road, Stevenage, UK SG1 2NY
Received (in Cambridge, UK) 17th July 2003, Accepted 3rd September 2003
First published as an Advance Article on the web 16th September 2003
Model studies are presented that suggest the mechanism of
the lanthanide(III) salt catalysed mono acylation of symmet-
rical diols proceeds via chelation of the diol and the
anhydride to the lanthanide salt, followed by an ‘intra-
molecular’ acyl transfer.
soluble, and solid LnCl3 is always seen during the course of the
reaction. To rule out the possibility of heterogeneous catalysis at
the surface of the undissolved lanthanide(III) salt, we prepared a
saturated THF solution of CeCl3 (conc. 0.018 M, 5 mol% w.r.t.
diol)7 and filtered it to remove any undissolved CeCl3. This
solution catalysed the mono acylation of meso-hydrobenzoin‡
equally effectively, implying that the catalysis is homogeneous.
The absence of competing donor solvent would lead to a rate
acceleration if the substrates bind to the metal before reacting,
and in CH2Cl2 solvent, this has been observed.
If the Ln(III) cation catalyses the reaction by chelating the diol
before acylation, then bound substrate would be acylated at a
much greater rate than any non-chelated substrate. The
observation that C2 diols are acylated faster than meso-diols
supports a mechanism where the substrate diol binds to the
metal catalyst in a bidentate fashion. The chelation of a meso-
diol to the metal salt might result in an unfavourable eclipsing
interaction between the R groups which is not present in the case
of the C2 diol. This interaction would slow the rate of chelation
and hence the rate of acylation (Fig. 1).
We have recently reported that lanthanide salts LnX3 (X = Cl,
OTf) catalyse the selective mono acylation of symmetrical 1,2-,
1,3- and 1,4-diols with carboxylic acid anhydrides.1–3 These
reactions are all the more remarkable as catalyst loadings can be
as low as 1 mol% (YbCl3) or even 0.1 mol% (Yb(OTf)3), while
10 equiv. of anhydride may be used without any detrimental
effect on the selectivity of the mono acylation reaction. Previous
to our work, the state-of-the-art in diol mono protection
included cleavage of cyclic acetals4 and continuous extraction
procedures.5 Lewis acid catalysts from the d- and p-block are
non-selective for this acylation. Herein, we propose a possible
mechanistic pathway, and present studies using sterically
encumbered substrates and complexes that model proposed
intermediates, which provide support for this hypothesis.
A number of observations have lead us to postulate a
mechanism for this reaction: (1) The acylation is faster, but
equally as selective, in a non co-ordinating solvent such as
CH2Cl2, compared with the standard THF solvent.6 (2) meso-
Diols or cyclic cis-1,2-diols are acylated at a reduced rate
compared to their C2-symmetric or cyclic trans-1,2-counter-
parts. (3) The rate of bis acylation is extremely slow. In general,
reactions left for many hours after mono acylation is complete
show only minimal bis acylation ( < 5%).
To identify the coordination of substrates in solution, we
prepared solutions of meso-hydrobenzoin and its mono ester at
reaction concentrations in an NMR tube and added solid CeCl3
(10 mol%). Unfortunately, we were unable to detect, by
chemical shift changes, any co-ordination of substrate to CeCl3
1
using H or 13C NMR spectroscopy. However, addition of
1
soluble [Eu(tfc)3] (10 mol%), affords sizable H NMR reso-
nance shifts of the carbinol (CH(OH)) protons in both meso-
hydrobenzoin (from 4.85 ppm to 5.22 ppm) and the mono-
acetate (5.81 and 4.96 ppm to 6.55 and 5.42 ppm),6,7 indicating
that both the diol and the mono ester product can chelate. As a
catalyst, [Eu(tfc)]3 (10 mol%) does generate mono-acylated
product in 90% yield.§
Our proposed reaction sequence, Scheme 1, shows both a diol
and an anhydride co-ordinated to the lanthanide(III) salt, 1. By
conventional wisdom this diol is now less nucleophilic.
However, co-ordination of the anhydride means that in 1 the
‘intra-molecular’ transfer of an activated acyl group is facili-
tated. The resultant 7-member chelate 2 is less stable than the
5-membered chelate originally formed by the substrate diol, so
the mono-acylated product dissociates, to be replaced by diol,
thus regenerating 1.
The use of a b-diketonate was considered a suitable
anhydride mimic for a model study, since it is structurally
similar to an anhydride. Treatment of [Pr(thd)3], (thd =
tetramethylheptanedioate, [ButC(O)CHC(O)But]2) with an
equivalent of meso-hydrobenzoin, afforded a pale green solid
after isolation and recrystallisation from diethyl ether, which
This proposed mechanism invokes homogeneous catalysis by
the lanthanide(III) salt. However, the halides are only poorly
1
was characterised as 3, eqn. (1).† The H NMR spectrum of 3
reveals strongly paramagnetically shifted and broadened en-
antiotopic carbinol and hydroxyl protons (meff Pr3+ (f2) = 3.58),
providing strong evidence for the formation of a praseodymium
diol chelate complex in solution. This complex may be a valid
model of an intermediate 1.
Scheme 1 Mechanistic hypothesis.
† Electronic supplementary information (ESI) available: experimental
details. See http://www.rsc.org/suppdata/cc/b3/b308171k/
Fig. 1 Diol co-ordination.
2588
CHEM. COMMUN., 2003, 2588–2589
This journal is © The Royal Society of Chemistry 2003

View Article Online
single intramolecular H-bonding interaction must selectively
weaken this OH and could, in principle, explain the selectivity
for monoacylation. The other OH (O61) also forms a hydrogen
bond, but to the lattice diethyl ether (O70); H61…O70
1.918(13) Å, O62…O70 2.707(2) Å and < (O62 H62 O70)
160.5(3)°. This could suggest an alternative, complex transition
state in which the solvent intermolecular H-bonding controls the
nucleophilic attack. However, the selectivity of this reaction in
CH2Cl2, in which hydrogen bonding would be weaker, supports
our confidence in the intramolecular hydrogen bonding as a
controlling factor.
(1)
Praesodymium complexes (10 mol%) were shown to catalyse
the pivolyation (PrCl3), and acylation ([Pr(thd)3] and 3) of
meso-hydrobenzoin Scheme 2. This shows that 3 is an effective
precatalyst as it is clearly capable of generating a species that
can catalyse the acylation reaction to the same extent as the free
lanthanide salt. To provide a structural as well as functional
model of the proposed transition state in which both substrates
are coordinated, crystals of 3 suitable for X-ray structural
analysis were grown from diethyl ether.¶
The solid-state molecular structure of 3, Fig. 2, reveals a
monomeric 8-coordinate Pr(III) centre, in which the rare earth
ion is ligated by both O atoms of meso-hydrobenzoin, forming
an approximate dodecahedral cation geometry. This is a rare
example of a structurally characterised lanthanide(III) polyol
adduct.7 Importantly, the meso-hydrobenzoin phenyl sub-
stituents are eclipsed with respect to one another, with a dihedral
angle of 43.5°, implying a significant steric interaction in the
Analogously to the NMR solution study, the reaction of
[Pr(thd)3] with the monoacylated product 4 afforded a para-
1
magnetically shifted adduct analogous to 2, according to H
NMR spectroscopy, but only the two starting materials could
ever be isolated upon crystallisation from a range of solvents.
The ‘transiency’ of the Pr-ester adduct supports the proposed
rapid replacement of product by diol substrate. If the chelated
substrate is acylated at a much greater rate than any non-
chelated substrate, this observation would account for both the
rate enhancement and the mono-selectivity seen.
To conclude, we believe that the bidentate coordination of
substrate diol to monoacylation catalysts of the form LnX3 is a
major factor in the control of selectively in diol desymmetrisa-
tion. We have shown that in solution both diol and monoester
product bind to the metal, but only the diol binds well enough to
form an isolable model complex. Structural characterisation of
a model intermediate shows a significant, asymmetric hydrogen
bond involving one of the diol hydroxyls.
We thank the EPSRC (P.L.A., Advanced Research Fellow-
ship, P.A.C., Grant GR/R22476/01), (L.S.N. Studentship fund-
ing), GSK (M.A.S., Industrial CASE studentship) and As-
traZeneca (P.A.C., unrestricted research support grant) for
financial support. The EPSRC National Mass. Spec. service,
Swansea, for Mass. Spec. determinations and Dr John Carey,
(GSK, Tonbridge) for a gift of Yb(OTf)3.
solid state. A weak chelation to the metal is evidenced by Cipso
–
OH bond lengths that are not significantly elongated upon
complexation. The HO–Pr and other distances are as antici-
pated.7
The meso-hydrobenzoin hydroxyl protons were located on
the difference fourier map. The O62 hydroxyl proton partici-
pates in a hydrogen-bonding interaction with O41 of the closest
b-diketonate ligand (close contacts are H62…O41 2.43(3)Å,
O62…O41 2.745(3) Å and < (O62–H61–O41) 104.2(2)°). This
Notes and references
‡ meso-Hydrobenzoin was used in all of our studies as ultimately we hope
to use the mechanistic insight gained to design asymmetric lanthanide (III
)
salts to use in desymmetrisation reactions.
§ We were unable to detect any enantioselectivity in the mono acylation of
meso-hydrobenzoin with acetic anhydride using Eu(tfc)3 as a catalyst.
¶ Characterising data for 3: isolated from hexane in 24% yield, 0.1226 g.
Anal. Calcd. for C47H71O8Pr: C, 62.25; H, 8.06. Found: C, 62.34; H, 8.17%.
Crystallographic data for 3: C51 H81 O9 Pr, fw 979.07, T = 150(2) K,
¯
triclinic, P1, a 10.576(3), b 12.662(4), c 20.988(6) Å, a 79.191(7), b
81.482(6), g 81.016(5)°. Dcalc 1.201 mg m23, mu 0.948 mm21, F(000)
1036, S 1.019, R1 0.0340, w(R)2 0.0828. CCDC 215708. See http://
www.rsc.org/suppdata/cc/b3/b308171k/ for crystallographic data in .cif or
other electronic format.
Scheme 2 Evaluation of praesodymium complexes.
1 P. A. Clarke, R. A. Holton and N. E. Kayaleh, Tetrahedron Lett., 2000,
41, 2687.
2 P. A. Clarke, N. E. Kayaleh, M. A. Smith, J. R. Baker, S. J. Bird and C.
Chan, J. Org. Chem., 2002, 67, 5226.
3 P. A. Clarke, Tetrahedron Lett., 2002, 43, 4761.
4 (a) M. Oikawa, A. Wada, F. Okazaki and S. Kusumoto, J. Org. Chem.,
1996, 61, 4469; (b) M. J. Martinelli, R. Vaidyanathan and V. V. Khau,
Tetrahedron Lett., 2000, 41, 3773; (c) N. Maezaki, A. Sakamoto, N.
Nagahashi, M. Soejima, Y.-X. Li, T. Imamura, N. Kojima, H. Ohishi, K-
i. Sakaguchi, C. Iwata and T. Tanaka, J. Org. Chem., 2000, 65, 3284; (d)
P. C. Zhu, J. Lin and C. U. Pittman, Jr., J. Org. Chem., 1995, 60, 5729;
(e) B. M. Choudary and P. N. Reddy, Synlett., 1995, 959; (f) M. Kinugasa,
T. Harada and A. Oku, J. Am. Chem. Soc., 1997, 119, 9067.
5 J. H. Babler and M. J. Coghlan, Tetrahedron Lett., 1979, 20, 1971.
6 L. S. Natrajan, J. J. Hall, A. J. Blake, C. Wilson and P. L. Arnold, J. Solid
State Chem, 2003, 171, 90.
Fig. 2 Thermal ellipsoid drawing of 3 (50% probability). CH3 groups, lattice
solvent and H atoms other than diol OH and OCH omitted for clarity.
Selected distances (Å) and angles (°): Pr–O(thd) range 2.368(2)–2.433(2),
Pr–O61 2.579(2), Pr–O62 2.651(2), O61–C61 1.442(3), O62–C62 1.439(3),
O61–Pr–O62 61.26(5), O53–Pr–O51 70.21(6).
7 Ch. Husson, P. Delangle, J. Pécaut and Ph. J. A. Vottéro, Inorg. Chem.,
1999, 38, 2012; P. Delangle, C. Husson, C. Lebrun, J. Pécaut and Ph. J.
A. Vottéro, Inorg. Chem., 2001, 40, 2953; S. J. Angyal and D. C. Craig,
Carbohydr. Res., 1993, 241, 1.
CHEM. COMMUN., 2003, 2588–2589
2589