The mechanism of the stereospecific intramolecular arylation of lithiated ureas: The role of Li+ probed by electronic structure calculations, and by NMR and IR spectroscopy

Article abstract of DOI:10.1002/ejoc.201101475

In situ NMR and IR spectroscopy studies were carried out on the rearrangement of lithiated N-benzyl-N'-aryl ureas, which involves N-to-C aryl transfer with retention of configuration. The IR spectroscopy studies revealed that initial benzylic lithiation w

Full text of DOI:10.1002/ejoc.201101475

FULL PAPER
DOI: 10.1002/ejoc.201101475
The Mechanism of the Stereospecific Intramolecular Arylation of Lithiated
Ureas: The Role of Li+ Probed by Electronic Structure Calculations, and by
NMR and IR Spectroscopy
Damian M. Grainger,[a] Alison Campbell Smith,[b] Mark A. Vincent,[a] Ian H. Hillier,*[a]
Andrew E. H. Wheatley,*[b] and Jonathan Clayden*[a]
Keywords: Reaction mechanisms / Rearrangement / Lithium / Lithiation / IR spectroscopy / Density functional calculations
In situ NMR and IR spectroscopy studies were carried out on
the rearrangement of lithiated N-benzyl-NЈ-aryl ureas, which
involves N-to-C aryl transfer with retention of configuration.
The IR spectroscopy studies revealed that initial benzylic li-
thiation was followed by migration of the aryl ring to yield
a lithiated urea product without a detectable dearomatised
intermediate. Similar results were obtained by NMR spec-
troscopy, but when 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-
pyrimidinone (DMPU) was added to the solvent mixture, a
transient dearomatised intermediate was detectable during
migration of a 1-naphthyl ring. DFT calculations highlight
the importance of coordinated lithium cations, and their mi-
to a site close the adjacent phenyl ring, allowing retentive
attack of the anionic centre on the more remote ring with
movement of the solvated lithium cation to the remote ring
stabilising the developing negative charge. A short-lived spi-
rocyclic intermediate is predicted to undergo elimination by
loss of the urea substituent, completing the migration. Co-
ordination of the carbonyl group to a second solvated lithium
cation appears to be essential for this step. Calculated shifts
for this intermediate when a 1-naphthyl ring is migrating are
consistent with the transient signals observed by NMR spec-
troscopy. Alternative pathways involving (1) invertive mi-
gration and (2) attack on the urea C=O group were also cal-
gration from one site to another, in the rearrangement. Re- culated and were found to require significantly higher en-
arrangement is initiated by migration of a solvated lithium
cation from the anionic centre of the starting organolithium
ergy transition states.
In complementary experimental work, we have shown
that, when carbamates carry an NЈ-aryl substituent, the
lithio derivatives (formed with either alkyllithiums or
Introduction
The construction of quaternary centres[1] adjacent to ni-
trogen[2–4] or oxygen[4,5] is a challenge that has been met in
a number of ways, primarily by the addition of nucleophiles
to one of the enantiotopic faces of a ketone or imine.[3–6]
Such an approach fails when the two substituents flanking
the ketone or imine are sterically and electronically similar,
and in these cases stereospecific construction of the quater-
nary centre from a pre-existing tertiary centre is an appeal-
ing alternative. Carbamate derivatives 1 of benzylic second-
ary amines[7,8] or alcohols[9,10] may be deprotonated to give
configurationally stable lithio derivatives 2, which react
with electrophiles either with retention or inversion,[11] al-
lowing the formation of quaternary centres in the products
3 (Scheme 1).[12,13] Aggarwal et al. have extended the utility
of the lithiated carbamates of secondary benzylic alcohols
by showing that they may be transformed into tertiary
alcohols and amines 4 through the addition and rearrange-
ment chemistry of boron derivatives.[14]
N-lithioamines) of benzylic ureas[15,16]
5 and carb-
amates[17,18] 6 undergo a remarkable NǞC aryl migration
(Scheme 2). Furthermore, related reactions occur with
allylic ureas,[19] cyclic ureas,[20] lithiated ureas generated by
carbolithiation,[21] and benzylic thiocarbamates.[22] Conver-
sion of the rearranged ureas 7 or carbamates 8 into the
amines 9 or alcohols 10 results in an overall stereospecific
arylation of an α-methylbenzylamine or α-methylbenzyl
alcohol derivative. However, the rearrangements of ureas
and carbamates differ from one another in one very impor-
tant respect: the aryl migration within the urea proceeds
with retention of stereochemistry at the migration ter-
minus,[15] whereas the aryl migration within the carbamate
proceeds with inversion of stereochemistry at the migration
terminus.[18] Chiral benzyllithium derivatives are known to
react with electrophiles with erratic stereospecificity (reten-
tive SE2ret or invertive SE2inv),[10–12] and while the detailed
stereochemical pathways by which such stereospecific elec-
trophilic substitution reactions proceed are yet to be fully
clarified, it is generally clear that Lewis base electrophiles
capable of coordination to lithium favour retentive over in-
vertive substitution.[12]
[a] School of Chemistry, University of Manchester,
Oxford Road, Manchester M13 9PL, UK
Fax: +44-161-275-4939
E-mail: clayden@man.ac.uk
[b] University Chemical Laboratory,
Lensfield Road, Cambridge CB2 1EW, UK
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
731

I. H. Hillier, A. E. H. Wheatley, J. Clayden et al.
FULL PAPER
Scheme 1. The synthesis of quaternary centres adjacent to N or O by lithiation of ureas and carbamates. For simplicity, external solvation
of Li+ is not detailed.
Scheme 2. The synthesis of quaternary centres adjacent to N or O by intramolecular aryl migration in lithiated ureas and carbamates.
For simplicity, external solvation of Li+ is not detailed.
Herein, we report electronic structure calculations and in reactive conformation 12Li cannot benefit from the intra-
situ IR and NMR spectroscopic studies aimed at the eluci- molecular Li–O coordination typical of lithiated carb-
dation of the pathway by which the 1,4-aryl migrations of amates,[23] and we assume that it is formed during the reac-
the lithiated ureas shown in Scheme 2 take place. In particu- tion pathway by C–N rotation from structure 5Li, intramo-
lar, the reason for the retention of stereochemistry during lecularly coordinated in the manner typically postulated for
migration in ureas is addressed.
such structures.[24]
The existence of the dearomatised spirocyclic intermedi-
ate 13cLi was confirmed by air oxidation of the reaction
mixture produced during migration of the 1-naphthyl ring
of 5c, from which the enone 16 was isolated.[15] However,
the general intermediacy of a dearomatised spirocycle 13Li
could not be proved, with attempts to trap similar struc-
tures during the migration of non-naphthyl rings being un-
successful.[25]
Results and Discussion
Experimental Evidence for the Mechanism of Aryl
Migration
Previous work showed unequivocally that the rearrange-
ment illustrated in Scheme 2 is an intramolecular pro-
cess.[15] Furthermore, evidence from the trapping of a de-
aromatised intermediate during migration of a naphthalene
ring[15] led us to propose that the rearrangement of lithiated
ureas proceeds through the pathway illustrated in Scheme 3,
NMR Spectroscopic Studies
The lithiation and rearrangement of 5 to 9 occurs in
in which nucleophilic attack of the organolithium centre in THF, but is accelerated considerably by the addition
the reactive conformation 12Li on the aryl ring generates of
1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone
the spirocyclic structure 13Li. For simplicity, ligated solvent (DMPU).[15] By carrying out the rearrangement in an
is not included in the structures. Subsequent elimination of NMR tube in the presence or absence of DMPU, we were
the lithiated urea substituent restores aromaticity and pro- able to elucidate some details of the structures of intermedi-
vides a product 14Li, which gives 9 upon protonation. The ates in the conversion of 5 to 9. These NMR spectroscopy
732
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 731–743

Intramolecular Arylation of Lithiated Ureas
Scheme 3. Schematic pathway for aryl migration. For simplicity, external solvation of Li+ is not detailed.
studies were carried out by using 5b and 5d as substrates
Ureas 5b (Figure 1) and 5d (Figure 2) were each dis-
to allow the two aryl rings of the substrate to be readily solved in THF in an NMR tube and cooled to –78 °C.
distinguished.
sBuLi was added and the mixture was agitated and immedi-
Figure 1. 1H NMR spectrum of 5b (a) with sBuLi in THF, –40 °C, assigned to the structure 5bLi and (b) with sBuLi and DMPU in
THF, –40 °C. Signals marked with an asterisk are assigned to a THF decomposition product.
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
733

I. H. Hillier, A. E. H. Wheatley, J. Clayden et al.
FULL PAPER
1
Figure 2. H NMR spectrum of 5d (a) with sBuLi in THF, DMPU, –40 °C, presumed to arise from the structure 14bLi; (b) with sBuLi
and DMPU in THF, –40 °C; and (c) with sBuLi and DMPU in THF after warming to room temperature. Signals marked with an asterisk
are assigned to a THF decomposition product.
ately transferred to the NMR probe at –40 °C. Figure 1 (a) of signals at δ = 6.5–8.0 ppm. Slow warming of the reaction
shows a portion of the spectrum obtained when using 5b. mixture in the NMR tube led to disappearance of the
Characteristic signals shifted upfield from the aromatic re- sharpened signals (Figure 2c), leaving only the broad sig-
gion indicate that complete lithiation of the substrate 5b nals at δ = 6.5–8.0 ppm similar to those seen when using 5b.
has given benzyllithium 5Li.[26] Delocalisation of charge We assume that these broad signals arise from the lithiated
into the ring accounts for the chemical shifts of the products 14Li, which are probably aggregated in solution.
MeOC6H4 aryl protons in the region of δ = 6–7 ppm and The sharp signals in part b of Figure 2, on the other hand,
the partial double bond between the anionic centre and the must arise from an intermediate en route to 14Li, which we
ring results in doubling up of those signals. The signals be- presume to be 13dLi. Eleven signals in the region above δ
tween δ = 5.9 and 4.8 ppm marked with asterisks appear to = 6.0 ppm might be expected for this structure, although
arise from the in situ elimination of lithiated THF to yield individual assignments, except for the remarkably deshield-
lithium but-3-ene 1-oxide, which is not the usual THF de- ed multiplet at δ = 9.5 ppm (see below), could not be made.
composition pathway,[27] but one that has previously been
observed in the presence of electron-rich ligands [hexameth-
ylphosphoramide (HMPA) and DMPU].[12,28]
In-situ IR Spectroscopy
Carrying out the same procedure with 5b in an NMR
tube, but now in the presence of DMPU (6 equiv.) gave a
The rearrangements of 5a and 5c were followed by in situ
very different spectrum, as shown in Figure 1 (b). Immedi- IR spectroscopy in THF. A mixture of 5a or 5c and THF
ately after transfer of the mixture to the NMR probe, broad in a three-necked flask equipped with a ReactIR probe was
signals were obtained, mainly in the aromatic region (δ = cooled to –60 °C. Once the temperature had stabilized,
6.5–7.5 ppm). Clearly, no 5bLi remains in the mixture, but spectroscopic analysis was commenced at a rate of one scan
there is also no evidence for dearomatised structure 13bLi. every 15 s. The initial spectrum was recorded, before sBuLi
The spectra obtained from 5d are more informative. Fig- (1.43 mL, 1.4 m in cyclohexane, 2 mmol) was added. The
ure 2 (a) illustrates the result of lithiating 5d in the absence reaction mixture was stirred at this temperature until the
of DMPU: as with 5b, the upfield shift out of the aromatic spectrum had stabilized and was then warmed in stages to
region indicated complete benzylic lithiation. On lithiation –40, 0 and finally to 25 °C; at each stage the spectrum was
in THF/DMPU – see part b of Figure 2 – a total of eleven allowed to stabilize if necessary before further warming.
new sharp signals immediately become evident between δ = Once a stable spectrum was observed at 25 °C, MeOH was
6.2 and 9.5 ppm, superimposed upon a much broader set added and a final spectrum was collected.
734
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 731–743

Intramolecular Arylation of Lithiated Ureas
Figures 3, 4 (for the rearrangement of 5a), 5 and 6 (for approximately at 1605 cm–1, which we consider arises from
the rearrangement of 5c) show the characteristic regions of the initial lithiated species 5aLi and 5cLi, or 12aLi and
the spectra where changes were observed. Both ureas show 12cLi. Recently reported in situ IR studies of the lithiation
an initial strong signal at 1645 cm–1 (blue lines in Figures 3 of a carbamate directly observed the “pre-lithiation com-
and 5) characteristic of a urea. In each case, the addition plex” between the substrate and lithiating agent,[29] the exis-
of sBuLi causes a shift of this carbonyl absorption to give tence of which had previously been deduced by kinetic stud-
a double peak (green line in Figures 3 and 5) centred ies on a related compound.[30] However, with ureas 5a and
Figure 3. Characteristic changes in the 1700–1500 cm–1 region of the IR spectrum during rearrangement of 5a. Four spectra taken at
different stages in the experiment are overlaid, with the time elapsed at the point of observation of each spectrum given (h:min:s). Blue
(00:00:53): urea prior to lithiation; green (00:07:28): after addition of sBuLi at –60 °C; red (01:44:14): after warming to 0 °C; pink
(02:32:23): immediately after addition of MeOH at 25 °C.
Figure 4. Perspective view of the changes in the IR spectrum during the rearrangement of 5a.
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
735

I. H. Hillier, A. E. H. Wheatley, J. Clayden et al.
FULL PAPER
5c we were unable to detect any transient intermediates be- the reactions were quenched with MeOH, and we assume
tween 5 and 5Li.
that the strong signals in the region of 1350 and 1575 cm–1
Warming both samples led to a further significant arise from C–O and C–N stretches within the lithio urea
change in the region illustrated in Figures 3 and 5, with function of rearranged compounds 14aLi and 14cLi. De-
a new strong signal forming close to 1575 cm–1 (red line), spite the evidence described above and that reported pre-
accompanied by a signal close to 1350 cm–1 (not illus- viously[15] for the intermediacy of a dearomatised interme-
trated). No further change was seen in either spectrum until diate 13Li in the rearrangement of 5d, no other intermedi-
Figure 5. Characteristic changes in the 1700–1500 cm–1 region of the IR spectrum during rearrangement of 5c. Four spectra taken at
different stages in the experiment are overlaid, with the time elapsed at the point of observation of each spectrum given (h:min:s). Blue
(00:03:49): urea prior to lithiation; green (00:07:49): after addition of sBuLi at –60 °C; red (01:24:54): after warming to 0 °C; pink
(02:31:54): immediately after addition of MeOH at 25 °C.
Figure 6. Perspective view of the changes in the IR spectrum during the rearrangement of 5c.
736
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 731–743

Intramolecular Arylation of Lithiated Ureas
ate was detectable in the IR studies. This difference may be
As an aid to the assignment of the in situ 1H NMR spec-
1
due to the use of DMPU to increase the rate of the reaction tra reported above, H chemical shifts were determined by
in the NMR spectroscopy studies – it is possible that inter- the gauge-independent atomic orbital (GIAO) approach,[33]
mediate 13Li is an observable intermediate only in the pres- using the 6-31++G(d,p) basis. All calculations were carried
ence of DMPU. Figures 4 and 6 illustrate the changes in out by using the Gaussian 03 program.[34]
the spectrum at each stage in the reaction.
Computational Results
Computational Studies
To gain some initial insights into various aspects of the
mechanism, we employed a small model (M1) with meth-
yllithium as the base, solvated by two THF molecules, to
give a tetrahedral Li+ ion when coordinated to the carbonyl
oxygen. The Li+ ion coordinated to the anionic centre is
solvated by a single THF molecule. The lowest energy ini-
tial reactant structure located (R1; Figure 7, a) had this Li+
close to the carbanion centre (2.09 Å, Table 1). To study the
1,4-aryl shift, we determined the TS structure (TS1) using
this model. The reaction involves the migration of the nega-
tive charge from the anion centre in the intermediate 12Li
to the oxygen (O–) centre of the product 14Li and, as ex-
pected, the solvated Li+ cation facilitates this charge trans-
To illuminate further the pathway by which aryl mi-
gration occurs in these systems, and the reasons for its re-
markable favourability, we have used electronic structure
calculations to explore the mechanism outlined in
Scheme 3, in which the lithiated carbanion 12Li undergoes
a 1,4-aryl shift to form 14Li where the negative charge is
located, at least formally, at oxygen. Such charge migration
is expected to be facilitated by lithium cations, which are
thus expected to undergo considerable movement during
the course of the reaction. We also studied the correspond-
ing 1,2-acyl shift reaction to give 15Li with a view to under-
standing why this usually favourable reaction does not oc-
cur.
Computational Details
Our computational strategy was to explore the mecha-
nism whereby the 1,4-aryl transfer occurs by employing
DFT calculations of realistic clusters, involving the reactant
5 in the form of its benzylic anion interacting with species
constructed from Li+, THF and a strong base to model
the environment during rearrangement. We have employed
several different models to try to understand the mecha-
nism, given that there is considerable uncertainty in both
the position and degree of coordination of the Li+ species.
Structures were optimized at the B3LYP/6-31G and B3LYP/
6-31++G** levels to obtain properly characterized minima
and transition-state (TS) structures. To study the potential
energy surfaces associated with these different models, our
strategy was to locate the appropriate TS structures and
then to follow each structure back and forward to obtain
the corresponding reactant and product structures.
Harmonic frequencies were computed at the B3LYP/6-
31G level and the resulting thermodynamic corrections (in-
cluding zero-point effects) were combined with the elec-
tronic energies determined with the larger basis set to yield
free energies. We have also examined the effect of using one
of the newer functionals designed by Truhlar and Zhao.[31]
Thus, we used the B3LYP/6-31++G** optimal structures to
determine single-point energies at the M06-2X /6-31++G**
level and corrected these to give free energies as before. As
far as bulk solvation is concerned, we decided, following
the work of Streitwieser and co-workers,[32] to ignore bulk
solvation effects. They found that the integral equation pol-
arisable continuum model (IEPCM) was poor at predicting
the solvation energies of a range of organic species in THF,
whilst with no bulk solvation the pK(Li) values of a range
of lithium compounds can be well predicted.
Figure 7. Stationary structures from model M1: (a) lowest energy
reactant (R1); (b) TS structure for retention of configuration (TS1);
(c) intermediate (I1); and (d) higher energy reactant (R2), leading
to retention of configuration.
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
737

I. H. Hillier, A. E. H. Wheatley, J. Clayden et al.
FULL PAPER
Table 1. Interatomic distances [Å] in stationary structures.[a]
Model
M1
Stereochemistry
retention
Structure
C–···Cphen
Li+···C–
Li+···Phen
N···Cphen
R1
R2
TS1
I1
TS2
R2
TS1
I1
TS2
R2
TS1
I1
R1
TS1
P
3.21
3.18
2.17
1.58
1.53
2.92
2.12
1.62
1.57
2.97
2.02
1.61
3.04
1.88
–
2.09
3.56
3.33
3.53
3.43
4.09
3.87
3.81
3.81
5.48
5.26
4.46
2.27
2.80
–
3.46
2.30
1.39
1.40
1.46
1.52
2.35
1.43
1.47
1.53
2.11
1.43
1.48
1.55
1.44
1.52
–
2.72 (L), 2.80 (A)
2.32
2.34
2.55
3.09 (L), 3.45 (A)
2.51
2.50
3.31
3.50 (L), 4.98 (A)
3.57
3.13
3.42
–
3.05
M2
retention
M3
M2
M3
retention
inversion
inversion
R1
TS1
P
2.90
1.62
1.56
4.00
4.97
4.32
1.44
1.77
3.27
3.45
Li+ Bonded to N–
[a] Cphen is the carbon atom that is attacked by C– of the carbanion centre, N is the nitrogen atom not bound to C–, and Li+ is the labile
lithium ion. Phen indicates one or other of the phenyl rings. Li+···Phen are average distances of the lithium ion to the carbon atoms of
the phenyl ring. In some TS structures the lithium cation is passed from one phenyl ring to the other; L and A label the ring the cation
leaves and the one it arrives at, respectively.
fer, as shown by its considerable movement during the invertive pathway. However, stereochemical integrity at the
course of the reaction. TS1, which leads to the experimen- carbanion centre is retained because conjugation with the
tally observed retention of configuration at the asymmetric adjacent phenyl group gives rise to a partial double bond,
centre, is shown in part b of Figure 7. In this structure, which prevents loss of stereospecificity by rotation or inver-
monosolvated Li+ has moved away from the carbanion cen- sion.
tre, by 1.24 Å relative to its position in R1, and now resides
Table 2. Free energies [kJmol–1], relative to the lowest energy struc-
ture (R1), of stationary structures on the potential energy surface
for retention and inversion of configuration.
between the two aromatic rings, at a distance of 2.7–2.8 Å
from them (Figure 7, b, and Table 1), thus stabilizing the
developing negative charge on the organic moiety of 13Li.
TS1 now leads to a five-membered-ring intermediate (I1,
Figure 7, c), corresponding to 13Li, which undergoes ring
opening to complete the 1,4-aryl shift. In preliminary calcu-
lations involving a migrating naphthyl ring, we find a sim-
ilar cyclic intermediate such as that found experimentally.
Methyllithium now stabilizes the negative charge that de-
velops on the carbonyl oxygen atom during the ring open-
ing of the cyclic intermediate I1, and it was found to be
essential to bring about completion of the reaction. Indeed,
no TS could be located for ring opening without coordina-
tion of MeLi to the carbonyl group. Consistent with this
observation are experimental observations that suggest
yields from the reaction are highest when two equivalents of
base are used. Nonetheless, this second equivalent of base is
not consumed during the reaction and the IR experiments
clearly show that only one equivalent of base is needed for
complete consumption of starting material.
Model
M1
Stereochemistry
retention
Structure
Free energy[a]
R1
R2
TS1
I1
TS2
P
0.0
14.6
44.8
–30.5
23.5
M2
retention
R2
TS1
I1
TS2
R2
TS1
I1
R1
TS1
P
R1
TS1
P
18.0 (24.8)
51.9 (52.0)
–3.1 (–29.0)
30.9 (30.7)
18.8 (51.8)
61.2 (82.9)
48.8 (31.3)
0.0 (0.0)
M3
M2
M3
retention
inversion
inversion
84.5 (70.3)
0.0 (0.0)
79.4 (92.1)
–61.4 (–66.3)
When TS1 was followed back to the reactant stage, the
lowest energy structure (R1) was not obtained. Instead
structure R2 (Figure 7d) was achieved, which was higher in
energy than R1 by 14.6 kJmol–1 (Table 2). In structure R2,
[a] The values are for the B3LYP functional; those for M06-2X are
given in parentheses.
The major conclusions from model M1 concerning the
the Li+ cation is located further from the carbanion centre 1,4-aryl shift relate to the movement of solvated Li+ during
than in R1 and is now located close to the adjacent phenyl the course of the reaction. We have found that (1) the reac-
ring (2.3 Å) (Figure 7, d, and Table 1). We note that the tive structure (R2) of 12Li has the solvated Li+ positioned
movement of Li+ away from the stereogenic centre in R2 over the phenyl ring adjacent to the carbanion centre rather
could potentially give rise to a stereochemically random or than over the carbanion centre itself where it is in the lower
738
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 731–743

Intramolecular Arylation of Lithiated Ureas
energy reactant structure (R1); (2) TS1, which leads to the Li+ coordinated to the carbanion centre was solvated by
five-membered cyclic intermediate 13Li has the solvated Li+ two and three THF molecules in models M2 and M3,
located between the two aromatic rings; and (3) coordina- respectively.
tion of lithium to the carbonyl oxygen atom is essential for
ring opening of 13Li to give the final product 14Li.
We now describe the use of models M2 and M3 to probe
both the observed reaction, leading to retention of stereo-
Having studied simple model M1, we next explored the chemistry, and also an alternative mechanism, leading to
use of larger models (M2, M3), which involve the base lith- stereochemical inversion. We should comment here that, in
ium diisopropylamide (LDA) in place of MeLi and more all of the models proposed, we have taken the reactive struc-
realistic solvation of Li+. LDA has been used experimen- ture to be 12Li, in which Li+ coordinates only to the carb-
tally as an alternative to alkyllithiums in rearrangements of anion centre, rather than structure 5Li, in which it is ad-
more sensitive ureas,[16,19] and is essential for stereospecific ditionally coordinated to the carbonyl oxygen. We find that,
rearrangements of the corresponding 1,4-aryl shift in carb- for model M2, structure 12Li is indeed of lower energy (by
amates.[17,18] We note that a crystal structure of an LDA– 18 kJmol–1) than 5Li, whereas if the LDA base is omitted
THF complex[35] shows a dimer with each Li+ ion coordi- from the model, then 5Li is preferred. This observation may
nated to a single THF molecule and sharing two amide li- underlie the fact that excess base facilitates the reaction.
gands with the second lithium ion. In the system studied
We consider first the stereochemically retentive mecha-
herein, the carbonyl oxygen replaces one of the THF mole- nism we studied using small model (M1) calculations. For
cules of this dimer. Our models involve one explicit LDA the larger systems, the overall picture leading to the forma-
dimer coordinated to the carbonyl oxygen, with the as- tion of the five-membered intermediate 13Li is nevertheless
sumption that a further molecule of base leads to deproton- similar to that found for M1. Thus, for both M2 (with two
ation of the urea.
solvating THF molecules) and M3 (with three solvating
Despite the presence of LDA dimer coordinated to the THF molecules) a reactant structure (e.g., R2, Figure 8a) is
carbonyl oxygen, the lithium cation associated with the determined. This structure is higher in energy by 18–
carbanion centre is still expected to interact with the car- 19 kJmol–1 (Table 2) than the lowest energy structure [e.g.,
bonyl oxygen. Thus, in models M2 and M3 the LDA dimer R1; Figure 8 (b)], with the solvated Li+ cation having
is coordinated through one of its lithium cations to the car- moved from being close to the carbanion centre (2.27 Å in
bonyl oxygen and the third Li+ is coordinated to the carb- R1) to a position near the adjacent phenyl ring (4.09 Å
anion centre and to other electron-rich moieties of the urea. from the carbanion centre in R2). In the case of the smaller
Figure 8. Stationary structures from model M2: (a) reactant (R2), leading to retention of configuration; (b) lower energy reactant (R1),
leading to inversion of configuration; and (c) TS structure (TS1), leading to retention of configuration. For clarity, the LDA dimer is
omitted.
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
739

I. H. Hillier, A. E. H. Wheatley, J. Clayden et al.
FULL PAPER
model (M1), the solvated Li+ in R1 is situated between the pathway, the charge developing on the migrating ring can-
phenyl rings [Figure 7 (d)]. However, the size of the solvated not therefore be stabilized by solvated Li+. We may thus
cation in both M2 and M3 prevents a similar ring-stacked attribute our finding that retention of stereochemistry is
structure; the rings are displaced with respect to each other preferred to these different degrees of TS stabilization.
[e.g., Figure 8 (b)]. Thus, there needs to be considerably Thus, we find that, for both M2 and M3, the absolute en-
more motion of the solvated Li+ to achieve the reactive ergy of the TS for inversion is greater than that for retention
structure (R2) than was necessary for model M1. The major of configuration by 33 and 18 kJmol–1, respectively. We
differences in the TS structures, leading to the formation of also noted that, for the reaction leading to inversion, we
the five-membered intermediate (I1), seen by comparing were unable to locate an anionic five-membered-ring inter-
TS1 for model M1 [Figure 7 (b)] with that for models M2 mediate, probably again due to the lack of stabilization by a
and M3 [Table 1 and, e.g., Figure 8 (c)], arise from the in- nearby lithium species. We previously reported preliminary
crease in the number of THF molecules that solvate the calculations of the corresponding reaction of a carbamate
moving Li+ cation. Thus, in all of the TS structures [TS1; 6 (Scheme 2) by employing a small model similar to M1.[17]
Figures 7 (b) and 8 (c)], the solvated Li+ cation is located We found that in this case inversion is favoured; a feature
between the two phenyl rings, but increased solvation in we attribute to the preferred binding of Li+ to the carb-
models M2 and M3 resulted in an increase in the distances amate’s benzylic oxygen atom rather than to the phenyl
of the Li+ ion from the phenyl rings (Table 1). This gives ring, as found herein for the urea.
rise to an increase in the height of the barrier, which leads
We have also investigated the corresponding 1,2-acyl
to I1, from 34 to 42 kJmol–1 for M2 and M3, respectively shift reaction to give 15, in which the anionic centre attacks
(Table 2), as the distance of the Li+ ion from the phenyl the carbonyl group. Reactions analogous to this are well
rings increases, and the associated charge stabilization de- known for related carbamates, amides, and allylic
creases. Steric factors, which are included in our models, ureas.[8,19,36] We employed model M2 and found that the TS
are also expected to play a role. We also note that there is (Figure 10) connected to a reactant structure higher in en-
a corresponding destabilization of the intermediate I1, as ergy (by 43.3 kJmol–1) than R1. This TS is higher in energy
the degree of solvation of the Li+ ion increases.
than R1 by 102.6 kJmol–1, and is thus unfavourable com-
We turn now to an alternative mechanism, leading to in- pared with both 1,4-aryl shift reactions. We may attribute
version of stereochemistry, which we have studied for mod- this to the considerable strain in the TS, which leads to a
els M2 and M3. The TS structure for this pathway is shown three-membered-ring intermediate.
in Figure 9. This TS structure is connected to the lowest
energy minimum (R1), so that for both M2 and M3 this
structure gives rise to an invertive rearrangement. In these
reactant structures both Li+(THF)2 and Li+(THF)3 ions are
in a relatively uncrowded region of space compared with
those leading to the retentive pathway. They thus lead to
TS structures in which these Li+(THF)n ions occupy the
face of the adjacent benzyl ring anti to the migrating ring
(Figure 9). In contrast to the stereochemically retentive
Figure 10. TS structure for attack at the carbonyl group and a 1,2
acyl shift (model M2). For clarity, the LDA dimer is omitted.
We see that the major features of the potential energy
surface for both retention and inversion of stereochemistry
are unchanged when the M06–2X functional is employed
(Table 2), although the actual magnitudes of the barriers
have altered somewhat. Thus, retention of stereochemistry
is again favoured over inversion and the calculated barriers
are larger for the more highly solvated cation.
Figure 9. TS structure for inversion (model M2). For clarity the
LDA dimer is omitted.
740
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 731–743

Intramolecular Arylation of Lithiated Ureas
Calculation of NMR Shifts
then the experimental shift of 50 cm–1 fits best with the li-
thiated species being R1 (12Li), which has a predicted shift
of 52 cm–1, rather than 5Li or R2 (12Li), which show calcu-
lated shifts of 15 and 27 cm–1 respectively. We also note
that, as far as using a comparison between experiment and
theory to identify whether R1 or R2 is observed experimen-
tally, although the evidence is not strong, the computed rel-
ative intensities of the two bands calculated at 1580–
1650 cm–1 for which that due to R2 is considerably more
intense than that due to R1, may suggest that reactant R1
is observed, in line with our conclusions from the calculated
C=O shifts. The spectrum measured for a reaction time of
about 2 h shows a broad band centred at 1575 cm–1, with
only an extremely weak feature at 1650 cm–1. This spectrum
correlates very well with our predicted spectrum of the
product anion (14Li) in which phenyl ring transfer has oc-
curred. Here the broad band is mainly due to the conju-
gated N=C–O– group, but also involves a number of C–H
modes that result in the observed broad band. Finally, the
very weak band observed at about 1650 cm–1 is calculated
to arise from stretching modes of the aromatic rings. We
have calculated the IR spectrum of the I1 in our predicted
reaction scheme; this structure shows an intense C=O
stretch at 1668 cm–1, which is completely absent in the mea-
sured spectrum. Thus, we conclude that this cyclised inter-
mediate has an extremely short lifetime.
For comparison with the spectrum of the transient spe-
cies observed by NMR spectroscopy during the rearrange-
ment of 5d [Figure 2 (b), assigned to structure 13dLi], we
used model M1 to calculate the 1H shifts for the cyclic inter-
mediate 13cLi, which differed from 13dLi by the lack of
a methoxy group. The computed chemical shifts shown in
Figure 11 contains 12 resonances in the aromatic region,
which yields 10 signals due to the averaging of the chemi-
cally equivalent pairs of hydrogen atoms at the ortho and
meta positions of the benzylic phenyl ring. We see that all
of the predicted resonances, except two, lie in the region δ
= 5.4–8.5 ppm, and thus, can be correlated with the group
of signals observed experimentally at δ = 5.0–7.8 ppm. Ex-
perimentally a doublet is observed at remarkably low field,
δ = 9.4 ppm. We assign this signal to the deshielded proton
at the peri position of the naphthyl ring, for which we also
predict a shift of δ = 9.37 ppm. Thus, since the experimental
spectrum is probably due to a mixture of species, agreement
between theory and experiment is satisfactory.
Table 3. Predicted peaks in IR spectra for Model M3.
Species
Mode
Frequency [cm–1
]
Intensity
Figure 11. Calculated proton chemical shifts [ppm] of intermediate
13cLi.
Urea (5)
C=O stretch
ring modes
C=O stretch
ring modes
C=O stretch
ring modes
C=O stretch
ring modes
C=O stretch
1632
1642–1664
1617
1637–1663
1580
1637–1657
1605
1635–1660
1668
1575, 1448, 1368 233, 168, 171
1636–1659 1–10
230
3–94
380
23–185
452
11–319
945
Li-Urea (5Li)
R1 (12Li)
R2 (12Li)
I1 (13Li)
Calculation of IR Spectra of Structures Along the Reaction
Pathway
273–28
814
We attempted to correlate the changes in the IR spectra
illustrated in Figures 3, 4, 5 and 6 with the minimum energy
structures computed for the intermediates in model M3.
Figures 3 and 5 show considerable changes in the experi-
mental spectra when sBuLi is added to solutions of ureas
5a and 5c in THF, respectively. In the dissolved urea 5a, an
intense band is observed at 1650 cm–1, which we associate
with the C=O stretch calculated to occur at 1632 cm–1
(Table 3); the weaker peaks seen at 1590–1600 cm–1 are as-
signed to aromatic ring motions calculated at 1642–
1664 cm–1. On lithiation, the C=O peak at 1650 cm–1 disap-
pears and a fairly intense broad band comprised of at least
Product (14Li) N=C(–O)–N modes
phenyl ring stretches
We note that our experimental IR spectra for the corre-
sponding naphthyl urea (5c) (Figure 5) are very similar to
those for 5a (Figure 3).
Conclusions
In this work we have sought reasons for why an organo-
two peaks appears at 1590–1610 cm–1 (Figure 5). This could lithium reagent might attack a phenyl ring intramolecularly
be due to either structures 5Li or 12Li (R1, R2). Our calcu- rather than undergo more conventional nucleophilic ad-
lations show a progressive decrease in the C=O frequency dition to a carbonyl group. We found that the degree of
from 5a to 5Li and 12Li (Table 3). Although inaccuracies coordination and movement of Li+ associated with the
in the calculated frequencies and the use of LDA in the carbanionic centre was critical to understanding the mecha-
calculations (as opposed to sBuLi in the IR spectroscopic nism of the 1,4-aryl shift. The major difference between the
work) make a definitive assignment of the initial lithiated structures, which led to retention or inversion of stereo-
species problematic, the calculated shifts in the C=O fre- chemistry, lay in the position of this solvated Li+ ion. We
quency upon lithiation are probably a better guide. If we found that the TS arising from R2 and leading to retention
take the more intense peak to be due to the C=O stretch, of configuration had the solvated cation located between
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
741

I. H. Hillier, A. E. H. Wheatley, J. Clayden et al.
FULL PAPER
the phenyl rings, whereas in the TS leading to inversion, gested that, as expected, restoration of aromaticity in a
which arose from R1, the cation was located on the benzylic naphthalene system was less favourable than in a simple
aryl ring and anti to the other phenyl ring. Thus, consider- phenyl ring.
able movement of the Li+ was required to attain the reactive
By means of NMR spectroscopy, we were able to follow
structure for retention of configuration. However, the TS the conversion of initial lithiated derivatives of benzylic
structure for stereochemically retentive reaction from R2 ureas 5b and 5d into species that could not be characterized,
was of lower energy than that from R1, which we attributed but the broad peaks of which were consistent with aggre-
to stabilization of the developing negative charge on the gated structures derived from 14Li. In the case of the
phenyl ring subject to nucleophilic attack. This TS struc- naphthyl migration in 5d, transient peaks formed that we
ture, which led to retention of stereochemistry, was also assigned to the dearomatized structure 13dLi. However, in
lower in energy than that leading to the alternative 1,2-acyl the case of phenyl migration in 5b no such intermediate was
shift reaction. Thus, we found that the surprising nucleo- observed.
philic attack at the aromatic ring was more favourable than
Finally, we note that, although we have located two alter-
the commonly observed attack at a carbonyl group. It native pathways leading to a 1,4-aryl shift in the ureas
would appear that the loss of aromaticity accompanying the studied herein, we believe that we cannot infer from these
1,4-aryl shift incurred a lower energetic penalty than the results the reasons for the stereochemical outcomes ob-
reaction that led to the 1,2-acyl shift via a strained TS state. served in the carbamates[17,18] or thiocarbamates;[22] each
Nucleophilic attack on aryl rings is not uncommon when system requires individual investigation.
the rings carry anion-stabilizing groups,[37] but is rarely ob-
served on rings as electron rich as those undergoing mi-
gration in the present rearrangements of lithiated ureas.
Experimental Section
These may carry up to three electron-donating groups.[15]
The “size” of the solvated Li+ complex, which had two
or three THF molecules associated with it, had important
implications for the structures of the reactants and TS
General: Starting materials were synthesised by the methods de-
scribed in ref.[15] In situ IR spectroscopy was used to follow the
rearrangements of 5a and 5c. A ReactIR 45m instrument was used
with an Evacuated Conduit arm equipped with a DiComp (Dia-
structure (TS1), leading to the formation of the five-mem-
mond) probe tip. The technique involves placing a probe into the
bered ring of 13Li. Thus, in the reactants (R1, R2), the
reaction mixture so that scans to establish IR absorbance can be
taken throughout the course of an experiment. Data was sub-
more highly solvated cation lay further from the anionic
centre, whilst in the TS structure (TS1), it was further from
sequently analyzed by using iC-IR 4.0 software. For these reactions
either of the two phenyl rings. In particular, in both R1 and
R2, there was increased steric crowding around the anionic
the change in absorbance of the characteristic carbonyl peak from
the urea group was followed because changes in this frequency al-
centre when the number of coordinated THF molecules in- lowed remote monitoring of other changes within the molecule.
creased from two to three, due to interaction of Li+(THF)3
and the benzylic and N-methyl substituents. Thus, in struc-
A three-necked flask was dried and subjected to a threefold evacua-
tion–nitrogen purge cycle to exclude oxygen and moisture. The Re-
ture R1, the more highly solvated lithium cation resided
over the adjacent phenyl ring, with its distance from the
anion centre increasing from 2.27 to 4.00 Å, in models M2
and M3, respectively, making the anion more available for
subsequent reaction. However, the corresponding barrier
for the reaction via TS1 was greater for model M3 than that
for M2, since the ability of the more highly solvated Li+ to
stabilize the developing negative charge on the migrating
phenyl was reduced. We previously found that, when Li+
was highly coordinated, there was a reduced tendency for it
to bind to an anion centre.[38] We found a similar effect
as far as the relative stability of the intermediate (I1) was
concerned. Thus, with increasing Li+ coordination the ef-
fective charge on the Li+ decreased, leading to reduced sta-
bilization of the pentadienyl anion of I1.
actIR probe was then inserted into the central neck, a temperature
probe into one of the side necks and the final neck used for ad-
dition of reagents and solvents. The urea reagent (5a: 0.537 g,
2 mmol or 5c: 0.637 g, 2 mmol) was added to the flask, dissolved
in THF (3 mL) and the reaction mixture was cooled to –60 °C.
Once this temperature had been achieved, IR scans were com-
menced at a rate of one scan every 15 s. The initial spectrum was
recorded and then sBuLi (1.43 mL, 1.4 m in cyclohexane, 2 mmol)
was added. The reaction mixture was stirred at this temperature
until the IR spectrum had stabilized and was then warmed in stages
to –40, 0 and finally 25 °C. At each stage the IR spectrum was
allowed to stabilize if necessary before further warming. Once a
stable spectrum was observed at 25 °C, MeOH (0.08 mL, 2 mmol)
was added and the spectrum was allowed to stabilize before ter-
mination of the experiment.
A comparison of the changes in IR spectra measured
during the course of reaction with the computed IR spectra
for the various minimum energy species, suggested that one
reactant structure (R1) was dominant and that the five-
membered intermediate had a short lifetime. Our calcula-
tions suggested that I1 (13Li) would probably not be ob-
served, with the barrier leading to ring opening to give 14Li
being only 34 kJmol–1 (model M2; Table 2). The trapping
Acknowledgments
We are grateful to the Engineering and Physical Sciences Research
Council (EPSRC) for support (grant numbers EP/G000816/1,
F069103 and S57785). Thanks also go to Mettler-Toledo for the
loan of a React IR 45m with an Evacuated Conduit arm and a
DiComp (Diamond) probe tip and to Drs. Jon Goode and Nigel
of the analogous structure in a naphthyl migration sug- Gaunt for technical assistance with the spectrometer.
742 Eur. J. Org. Chem. 2012, 731–743
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Intramolecular Arylation of Lithiated Ureas
[24] D. Hoppe, F. Marr, M. Brüggemann, in: Organolithiums in
Enantioselective Synthesis (Ed.: D. M. Hodgson), Springer,
Berlin, 2003.
[1]
[2]
E. J. Corey, A. Guzman-Perez, Angew. Chem. 1998, 110, 2092;
Angew. Chem. Int. Ed. 1998, 37, 388.
J. Clayden, M. Donnard, J. Lefranc, D. J. Tetlow, Chem. Com-
mun. 2011, 4624; E. M. Forbeck, C. D. Evans, J. A. Gilleran,
P. Li, M. Joullié, J. Am. Chem. Soc. 2007, 129, 14463; O. Toma-
shenko, V. Sokolov, A. Tomashevskiy, A. de Meijere, Synlett
2007, 652.
T. R. Ramadhar, R. A. Batey, Synthesis 2011, 1321.
M. Shibasaki, M. Kanai, Chem. Rev. 2008, 108, 2853; O. Riant,
J. Hannedouche, Org. Biomol. Chem. 2007, 5, 873; P. G. Cozzi,
R. Hilgraf, N. Zimmerman, Eur. J. Org. Chem. 2007, 5969.
M. Hatano, I. Kazuaki, Synthesis 2008, 1647.
[25] J. Clayden, R. A. Harvey, unpublished results.
[26] G. Fraenkel, K. V. Martin, J. Am. Chem. Soc. 1995, 117, 10336;
G. Fraenkel, J. H. Duncan, K. V. Martin, J. Wang, J. Am.
Chem. Soc. 1999, 121, 10538.
[27] R. E. Bates, L. M. Kroposki, D. E. Potter, J. Org. Chem. 1972,
37, 560.
[28] J. Clayden, S. A. Yasin, New J. Chem. 2002, 26, 191; I. Fleming,
S. R. Mack, B. P. Clark, Chem. Commun. 1998, 713.
[29] D. Stead, G. Carbone, P. O’Brien, K. R. Campos, I. Coldham,
A. Sanderson, J. Am. Chem. Soc. 2010, 132, 7260.
[3]
[4]
[5]
[6]
[30] D. J. Gallagher, P. Beak, J. Org. Chem. 1995, 60, 7092.
[31] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215.
[32] A. Streitwieser, J. R. Reyes, T. Singhapricha, S. Vu, R. Shah, J.
Org. Chem. 2010, 75, 3821.
R. Bloch, Chem. Rev. 1998, 98, 1407; P. I. Dosa, G. C. Fu, J.
Am. Chem. Soc. 1998, 120, 445.
[7]
N. C. Faibish, Y. S. Park, S. Lee, P. Beak, J. Am. Chem. Soc.
1997, 119, 11561.
O. Hara, M. Ito, Y. Hamada, Tetrahedron Lett. 1998, 39, 5537.
D. Hoppe, F. Hintze, P. Tebben, Angew. Chem. 1990, 102, 1455;
Angew. Chem. Int. Ed. Engl. 1990, 29, 1422; C. Derwing, D.
Hoppe, Synthesis 1996, 149; F. Hammerschmidt, A. Han-
ninger, Chem. Ber. 1995, 128, 1069.
A. Carstens, D. Hoppe, Tetrahedron 1994, 50, 6097.
R. E. Gawley, Tetrahedron Lett. 1999, 40, 4297; R. E. Gawley,
E. Low, Q. Zhang, R. Harrris, J. Am. Chem. Soc. 2000, 122,
3344.
[33] R. Ditchfield, Mol. Phys. 1974, 27, 789.
[8]
[9]
[34] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tom-
asi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratch-
ian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Sal-
vador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Ba-
boul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Lia-
shenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T.
Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, C. Gonzalez, J. A. Pople, Gaussian 03, rev. D.02,
Gaussian Inc., Wallingford, CT, 2004.
[10]
[11]
[12] J. Clayden, Organolithiums: Selectivity for Synthesis, Pergamon,
Oxford, 2002.
[13] Top. Organomet. Chem. Vol. 5 (Ed.: D. M. Hodgson), 2003.
[14] J. L. Stymiest, V. Bagutski, R. M. French, V. K. Aggarwal, Na-
ture 2008, 456, 778; V. Bagutski, R. M. French, V. K. Aggar-
wal, Angew. Chem. Int. Ed. 2010, 49, 5142; V. Bagutski, T. G.
Elford, V. K. Aggarwal, Angew. Chem. Int. Ed. 2011, 50, 1080.
[15] J. Clayden, J. Dufour, D. Grainger, M. Helliwell, J. Am. Chem.
Soc. 2007, 129, 7488.
[16] J. Clayden, U. Hennecke, Org. Lett. 2008, 10, 3567.
[17] J. Clayden, W. Farnaby, D. M. Grainger, U. Hennecke, M. [35] P. G. Williard, J. M. Salvino, J. Org. Chem. 1993, 58, 1.
Mancinelli, D. J. Tetlow, I. H. Hillier, M. A. Vincent, J. Am.
Chem. Soc. 2009, 131, 3410.
[18] A. M. Fournier, R. A. Brown, W. Farnaby, H. Miyatake-Ondo-
zabal, J. Clayden, Org. Lett. 2010, 12, 2222.
[36] S. Superchi, N. Sotomayor, G. Miao, J. Babu, M. G. Campbell,
V. Snieckus, Tetrahedron Lett. 1996, 37, 6061; B. Cottineau, I.
Gillaizeau, J. Farard, M.-C. Auclair, G. Coudert, Synlett 2007,
1925.
[19] D. J. Tetlow, U. Hennecke, J. Raftery, M. J. Waring, D. S.
Clarke, J. Clayden, Org. Lett. 2010, 12, 5442.
[20] R. Bach, J. Clayden, U. Hennecke, Synlett 2009, 421.
[21] J. Clayden, M. Donnard, J. Lefranc, A. Minassi, D. J. Tetlow,
J. Am. Chem. Soc. 2010, 132, 6624.
[22] P. MacLellan, J. Clayden, Chem. Commun. 2011, 3395.
[23] M. Marsch, K. Harms, O. Zschage, D. Hoppe, G. Boche, An-
gew. Chem. 1991, 103, 338; Angew. Chem. Int. Ed. Engl. 1991,
30, 321.
[37] C. Paradisi, in: Comprehensive Organic Synthesis, vol. 4 (Eds.:
C. H. Heathcock, I. Fleming), Pergamon, Oxford, 1991, p. 423.
[38] D. R. Armstrong, S. R. Boss, J. Clayden, R. Haigh, B. A. Kir-
mani, D. J. Linton, P. Schooler, A. E. H. Wheatley, Angew.
Chem. 2004, 116, 2187; Angew. Chem. Int. Ed. 2004, 43, 2135;
D. R. Armstrong, J. Clayden, D. J. Linton, P. Schooler,
A. E. H. Wheatley, Chem. Commun. 2003, 1694.
Received: October 11, 2011
Published Online: December 6, 2011
Eur. J. Org. Chem. 2012, 731–743
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
743