Asymmetric synthesis of trans-disubstituted cyclopropanes using phosphine oxides and phosphine boranes

Article abstract of DOI:10.1039/b817433d

The stereocontrolled synthesis of trans-disubstituted cyclopropylketones has been achieved from β-alkyl, γ-benzoyl phosphine oxides via a three-step cascade reaction incorporating an acyl transfer, phosphinoyl transfer and cyclisation to form the cyclopropane. Using Evans' chiral oxazolidinone auxiliary and by masking the phosphine oxide moiety as a phosphine borane we have extended the method to the synthesis of enantiomerically-enriched trans-disubstituted cyclopropyl ketones.

Full text of DOI:10.1039/b817433d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Asymmetric synthesis of trans-disubstituted cyclopropanes using phosphine
oxides and phosphine boranes†
Celia Clarke,a Ste´phanie Foussat,a David J. Fox,*b Daniel Sejer Pedersenc and Stuart Warrena
Received 3rd October 2008, Accepted 30th October 2008
First published as an Advance Article on the web 12th December 2008
DOI: 10.1039/b817433d
The stereocontrolled synthesis of trans-disubstituted cyclopropylketones has been achieved from
b-alkyl, g-benzoyl phosphine oxides via a three-step cascade reaction incorporating an acyl transfer,
phosphinoyl transfer and cyclisation to form the cyclopropane. Using Evans’ chiral oxazolidinone
auxiliary and by masking the phosphine oxide moiety as a phosphine borane we have extended the
method to the synthesis of enantiomerically-enriched trans-disubstituted cyclopropyl ketones.
The rigidity of the cyclopropane ring makes it an appealing
structural motif for the preparation of molecules with defined
orientation of pendant functional groups whilst maintaining the
hydrophobic character of linear alkyl chains. The stereocontrolled
synthesis of cyclopropanes has therefore been the centre of
much research effort.1 Cascade ring closing reactions involving
phosphorus transfer, to generate both nucleophile and leaving
group, have yielded cyclopropanes generally with high stere-
ospecificity and selectivity. Phosphine oxides,2–5 phosphonates,6–8
and phosphonium salts9 have been employed in this manner.
Studies on diphenylphosphine oxide mediated cyclopropanation
cascade reactions have shown that a variety of differentially
substituted cyclopropanes can be produced with excellent stere-
ocontrol (Scheme 1, eq. 1). We have achieved the synthesis of
di- and trisubstituted cyclopropanes in this manner by including
substituents at the a- and g-positions5,10 and also at the b-
and g-positions11,12 of the cyclisation substrates. Recently, we re-
ported the asymmetric synthesis of disubstituted trans-cyclopropyl
ketones3,13 and protected trans-cyclopropane amino acids14 using
g-substituted cyclopropane precursors (Scheme 1, eq. 1).
However, to date the effect of b-substituents has been inves-
tigated only in the presence of g-substituents.11,12 In order to
investigate the effect of substituents b to the phosphine oxide
group on the stereoselectivity of the cyclopropanation step, it
was necessary to find a method of introducing a substituent at
this carbon whilst keeping the a- and g-positions unsubstituted.
3-Diphenylphosphinoyl propionic acid derivatives 6 have a carbon
backbone of the required length whilst having a group on the b-
carbon capable of directing deprotonation and alkylation (7). Im-
Scheme 1 (i) LDA, THF, -78 to 0 ◦C, 95% (>95% ee)3 R1 = OR or chiral
portantly, this could provide a route to enantiomerically-enriched
auxiliary.
cyclopropane precursors (9) based on a chiral auxiliary strategy
(Scheme 1, eq. 2, R1 = auxiliary). We now wish to report our results
on the racemic synthesis of disubstituted trans-cyclopropanes
(10) utilising b-substituted phosphine oxides. Moreover, we have
extended this methodology to the enantioselective synthesis of
trans-cyclopropanes using phosphine boranes and a phenylalanine
based oxazolidinone auxiliary developed by Evans and Gage.15,16
We decided to study the cyclopropanation cascade reaction using
the b-unsubstituted phosphine oxide 21 and three b-substituted
phosphine oxides 22 to 24 (Scheme 2).
Diphenylmethyl phosphine oxide 11 was alkylated with tert-
butyl bromoacetate to give phosphinoyl ester 12. LHMDS in THF
is known not to lithiate diphenyl alkyl phosphine oxides,17 but does
aCambridge University, University Chemical Laboratory, Lensfield Road,
Cambridge, U.K. CB2 1EW
bDepartment of Chemistry, University of Warwick, Gibbet Hill Road,
Coventry, U.K. CV4 7AL. E-mail: d.j.fox@warwick.ac.uk
cUniversity of Copenhagen, Department of Medicinal Chemistry, Univer-
sitetsparken 2, 2100, Copenhagen, Denmark
† Electronic supplementary information (ESI) available: experimental and
analytical details for all compounds, and coordinates and energies of
calculated structures. See DOI: 10.1039/b817433d
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1323–1328 | 1323
©

View Article Online
Scheme 2 (i) n-BuLi, tert-butyl bromoacetate, THF, -78 ◦C to room temperature, 40%; (ii) LHMDS, THF, -78 ◦C to room temperature, 13 48%,
14 64%; (iii) LiAlH4, THF, 15 77%, 17 56%, 18 94%; (iv) 2-methyl-propenol, pyridine, 0 ◦C to reflux, 28%; (v) BH3, THF; H2O2, NaOH (aq), 68%;
(vi) PhCOCl, Et3N, DMAP, CH2Cl2, 21 76%, 22 91%, 23 84%, 24 42%; (vii) LDA, THF, -78 ◦C to room temperature, 25 77%, 27 54%, 28 76%, 29 99%;
(viii) t-BuOK, t-BuOH, 35 ◦C, 87%.
enolize standard esters. This reactivity allowed for the selective
alkylation of ester 12 adjacent to the carbonyl group giving
derivatives 13 and 14. Reduction of esters 12 to 14 with LiAlH4 and
benzoylation of the resulting alcohols gave the desired carboxylic
esters 21, 23 and 24. In the case of the methyl substituted phosphine
oxide 22 an alternative synthetic route was used starting from
chlorodiphenylphosphine 19 and 2-methyl-propenol producing
olefin 20 by an allylic Arbuzov rearrangement. Hydroboration
of olefin 20 followed by benzoylation of the resulting alcohol 16
gave the required carboxylic ester 22.
The four phosphine oxides were treated with the conditions
that we have developed for the phosphine oxide mediated cy-
clopropanation cascade reaction.3 The cyclopropanation results
highlight the importance of substitution for driving the cascade
reaction to completion: unlike the other substrates, unsubstituted
21 (Scheme 2) produces no cyclopropane under the standard
reaction conditions but only acyl transfer product 25. The Thorpe–
Ingold effect suggests that increased substitution on the forming
ring should encourage and stabilise ring closures (e.g. 30 to 31 and
32 to 33) (Scheme 3). Lithiation and ring closure of all substrates
occurs, but substitution at the b-position slows the ring opening of
the tetrahedral intermediate 31 compared to its rate of formation.
The alkoxy-keto-phosphine oxides 32 are significantly more acidic
than the non-keto phosphine oxide starting materials and so it
is likely that any ketone formed will be deprotonated by any
open-chain lithiated starting material 30 producing an inactive
enolate 34, incapable of further phosphinoyl transfer. Backbone
substitution should increase the rates of phosphinoyl transfer
and cyclopropanation reactions due to stabilisation of their cyclic
transition states. This is consistent with the increase in yield of
cyclopropanes 27 to 29 as the size of the substituent increases.
The acyl transfer product 25 could, however, be converted to the
desired mono-substituted cyclopropane 26 upon treatment with
t-BuOK in t-BuOH.
In every case, only trans-cyclopropane 10 was observed. Dia-
stereoselectivity induced by the b-substituent is in accordance with
the proposed transition state (Scheme 4) where steric interactions
are minimised in enolate 33 if the substituent R can lie trans to the
phenyl ketone during ring formation.
Scheme 4 (i) LDA, THF -78 ◦C to room temperature.
Having established that this approach is a viable route to trans-
cyclopropanes, we decided to investigate if the method could be
applied to the asymmetric synthesis of trans-cyclopropanes by
performing the required b-alkylations using the Evans oxazo-
lidinone auxiliary.15,16 The required oxazolidinone 35 was easily
prepared from phosphine oxide 12 in two steps by deprotection of
the tert-butyl ester followed by coupling with the chiral auxiliary
(Scheme 5). However, alkylation of oxazolidinone 35 with benzyl
bromide using LHMDS in THF to give phosphine oxide 36 was
Scheme 3
1324 | Org. Biomol. Chem., 2009, 7, 1323–1328
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 1 Alkylation of oxazolidinone 44
Entry
Base
RX
Product
Yield (%)a
1
2
3
4
LHMDS
LHMDS
LHMDS
NHMDS
MeI
CH2 CHCH2Br
BnBr
45
46
47
45
76
46
57
53
=
MeI
a Isolated yield of pure major diastereoisomer.
Scheme
5
(i) a. TFA, CH2Cl2, 99%; b. t-BuCOCl, Et3N, THF,
(S)-(-)-4-benzyl-2-oxazolidinone, n-BuLi, -78 ◦C to room temperature,
62%; (ii) LHMDS, THF, BnBr, 0%.
unsuccessful, producing only free auxiliary 37, alkylated auxiliary
38 and carboxylic acid 39.
Similar auxiliary alkylation products have been observed by
both Seebach and Hintermann18 and Davies et al.19 Both proposed
that the enolate fragments via a ketene decomposition pathway
to generate a lithiated auxiliary and that this is then alkylated
under the reaction conditions. Seebach and Hintermann improved
benzylation yields using sodium and zinc enolates.18 Attempted
alkylations of phosphine oxide 35 using either a potassium base
or by transmetallation with zinc, however, also failed suggesting
that enolate reactivity was not the problem. Phosphine oxides
are good Lewis bases and in the (Z)-enolate 40 (Scheme 6) the
phosphine oxide lies on a flexible side chain cis- to the exo-oxygen
and could possibly complex the lithium. If this coordination of
the phosphine oxide to the lithium is competitive with that of the
carbonyl of the auxiliary (40), then the lithium could be removed
from its conventional chelation site (41) making the auxiliary a
better leaving group than when coordinated in chelate 40.
Scheme 7 (i) a. BH3, THF, LiAlH4, THF; b. methyl acrylate, MeOH,
NaOMe, 0 ◦C, 96%; (ii) KOH, MeOH, H2O, 78%; (iii) pivaloyl chloride,
Et3N, THF, (S)-(-)-4-benzyl-2-oxazolidinone, n-BuLi, -78 ◦C to room
temperature, 71%; (iv) LHMDS or NHMDS, THF, RX, -78 ◦C to room
temperature. See Table 1 for yields.
products was assigned based on the model proposed by Evans and
co-workers.22
Small quantities of free auxiliary 37 and carboxylic acid 39
were present in the crude reaction mixtures. Even though their
diastereoselectivities, the alkylated products could be isolated as
single diastereoisomers after purification. Extending the reaction
time from 18 to 42 h at 0 ◦C resulted in a slight reduction in yield.
Using NHMDS (entry 4) instead of LHMDS also resulted in a
drop in yield. Next, alkylation products 45 to 47 were reduced by
the method of Roques and co-workers using sodium borohydride
and lithium chloride to give primary alcohols 48 to 50 in moderate
to excellent yield (Scheme 8).23
Scheme 6 Plausible mechanism for decomposition of enolate 40.
Successful alkylation of phosphine oxide 35 requires the 6-
membered chelate 40 to be more stable relative to the 7-membered
structure 41. This could possibly be achieved by reducing the
Lewis basicity of the phosphine oxide functionality. Phosphine
boranes would not be expected to chelate lithium as well as the
corresponding oxides. Moreover, phosphine boranes are easily
deprotected with trialkylamines under mild conditions and reoxi-
dised to the corresponding phosphine oxides.20 Hence we decided
to synthesise phosphine borane 44 (Scheme 7) as a protected form
of the required phosphine oxide.
Phosphine borane 42 was synthesised via reduction21 of
chlorodiphenylphosphine to diphenylphosphine borane, and con-
jugate addition to methyl acrylate. Hydrolysis of the carboxylic
ester and coupling to Evans’ auxiliary gave phosphine borane
44 in good yield. As anticipated, alkylation of phosphine borane
44 using LHMDS and a variety of electrophiles gave the desired
products 45 to 47 (Table 1). The stereochemistry of the alkylation
Scheme 8 (i) NaBH4, LiCl, EtOH, 48 99%, 49 57%, 50 95%; (ii) a.
DABCO, PhMe, 40 ◦C; b. H2O2; c. BzCl, Et3N, DMAP, 51 96%, 52 70%,
53 46%; (iii) LDA, THF, -78 to 0 ◦C, 54 60%, 55 77%, 56 72%.
Pellon and co-workers¢ method of phosphine decomplexation
employing DABCO in toluene20 was used with our substrates 48
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1323–1328 | 1325
©

View Article Online
to 50 and the resulting crude phosphines oxidised with hydrogen
peroxide to give the corresponding phosphine oxides that were
benzoylated, without prior isolation, to give carboxylic esters 51 to
53. Subjecting enantiomerically enriched cyclopropane precursors
51 to 53 to the standard cyclopropanation conditions completed
the three-step cascade reaction to give the desired cyclopropanes
54 to 56 in good yield. As previously observed for the racemic
cyclopropanation (Scheme 2), only trans-cyclopropane product
was isolated.
The significant drop in energy from Evans’ chelate 57 to seven-
membered chelate 58 indicates that the phosphine oxide is an
excellent ligand for lithium. A further set of structures in which
the auxiliary carbonyl, the enolate oxygen and the phosphine oxide
are all chelated to the lithium were also found. These are even
more stable (-48.5 to -44.5 kJ mol-1 compared to 57), and have
a structure where the p-systems of the auxiliary and the enolate
are not coplanar (59 in Fig. 1). The tilting of the auxiliary out of
the plane of the enolate may block the approach of an electrophile
from one side, while the phosphine oxide blocks the other side
(Fig. 1). The lack of experimentally observed reactivity of these
phosphine-oxide-containing enolates is therefore explained by
their propensity to form potentially unreactive chelates such as
structure 59.
A similar study shows that the three different classes of chelates
also exist for the phosphine borane enolates (Fig. 2). In this
case, however, the auxiliary–enolate chelates such as 60 (0 to
2.3 kJ mol-1) are more stable than the enolate–phosphine borane
chelates such as 61 (+20.8 to +21.1 kJ mol-1). This indicates that
the phosphine borane is a significantly worse ligand for lithium
than the phosphine oxide. The third class of chelate in which both
the auxiliary and the borane are chelated (e.g. 62) is more stable
than the corresponding Evans chelates (-13.4 to -8.7 kJ mol-1),
but this difference is much less than in the phosphine oxide series.
The phosphine borane chelates have similar structural features to
the phosphine oxides, but with the B–H bonds coordinating to the
lithium via agostic-type interactions which are much weaker than
the lone pair coordination of the phoshine oxide.
Molecular modelling
The high levels of stereoselectivity frequently observed in Evans’
enolate reactions are proposed to occur due to the significant
difference in steric hindrance of the approach of an electrophile to
each face of the enolate. The six-membered chelate in the enolate
controls the position of the orientation of the auxiliary relative
to the enolate, and may also stabilise the enolate by inhibiting the
departure of the auxiliary anion. DFT energy minimisations using
PC-GAMESS24 (B3LYP/6-31G(d))25–30 of simplified versions of
the phosphine oxide Evans lithium enolate 57 provided five stable
enolate monomers chelated by one dimethyl ether molecule with
a range of energies (with the lowest energy structure 57 arbitrarily
taken as zero) of 0 to 4.9 kJ mol-1 (see ESI for details†). Structures
in which the lithium is chelated between the enolate oxygen
and the phosphine oxide were also modelled and found to have
energies of -28.8 to -28.7 kJ mol-1 compared to Evans’ chelate
structure 57. The lowest energy conformation 58 is shown in Fig. 1.
Fig. 1
Fig. 2
1326 | Org. Biomol. Chem., 2009, 7, 1323–1328
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
It should be noted that no account of the change in conforma-
tional entropy has been made here, and no absolute equilibrium
positions have been calculated. The calculations, however, imply
that far less of the phosphine borane enolate will be trapped in an
unreactive chelate than the equivalent phosphine oxide, and this is
why the enolate alkylation occurs in the former case. In addition,
warming of the enolates may increase the proportion of less stable
chelates. For the phosphine oxide, the next most stable structure
lacks chelation of the auxiliary, and this may explain why the
enolate decomposes with loss of, and subsequent alkylation of, the
oxazolidinone. For the phosphine borane enolates, however, loss
of chelation to the auxiliary carbonyl is energetically unfavourable,
which is consistent with the increased observed stability, and
stereocontrolled alkylation, of the borane enolate.
In brief, we have demonstrated that b-substituted phosphine
oxides undergo the phosphine oxide mediated cyclopropanation
cascade reaction to produce trans-disubstituted cyclopropyl ke-
tones in good yield. Moreover, we have extended the method to
the enantioselective synthesis of cyclopropyl ketones by employing
Evans’ oxazolidinone auxiliary and masking the phosphine oxide
as a phosphine borane.
29.0 (d, J 36, PCH2) and 20.9 (d, J 11.5, Me); dP (162 MHz; CDCl3)
15.6–15.3 (m); m/z (ESI) 468 (65%, MNa+), 454 (51, MNa - BH3),
446 (33, MH) and 369 (100, MH - Ph) (Found: MNa+, 468.19000.
C26H29NO3PBNa requires M, 468.18758).
(R)-Diphenyl(3-hydroxy-2-methylpropyl)phosphine borane 48
By the method of Roques et al.,23 to a suspension of sodium
borohydride (27 mg, 0.72 mmol) and lithium chloride (31 mg,
0.72 mmol) in ethanol : THF (2.9 cm3 : 2.9 cm3), stirred at
0
◦C under nitrogen, was added a solution of (4S,2¢R)-3-[3¢-
(boronatodiphenylphosphinyl)-2-methylpropionyl)]-4-(phenylme-
thyl)oxazolidin-2-one 45 (80 mg, 0.18 mmol) in ethanol : THF
(0.72 cm3 : 0.72 cm3). The resulting mixture was stirred for 18 h,
allowing to warm to room temperature slowly. The mixture was
treated with acetone (1 cm3) and the solvents removed in vacuo.
The residue was partitioned between EtOAc (2 ¥ 25 cm3) and
water (15 cm3) and the combined organic layers dried (Na2SO4),
filtered and the solvents removed in vacuo to give an oil. The oil
was purified by flash column chromatography (SiO2, EtOAc–
hexane 1 : 1, v/v) to give alcohol 48 (50 mg, 99%) as an oil; Rf
0.3 (EtOAc–hexane, 1 : 1, v/v); [a]2D3 (c = 0.5, CHCl3) -4.7; IR
nmax (film)/cm-1 3372 (br, O–H), 2925 (C–H), 2381 (B–H), 1589
Experimental procedures
=
(C C, Ph) and 1436 (P–Ph); dH (500 MHz; CDCl3) 7.73–7.68
(4H, m, Ph), 7.49–7.40 (6H, m, Ph), 3.51 (1H, ddd, J 11, 5 and
1.5, CHAHBO), 3.40 (1H, dd, J 11 and 6, CHAHBO), 2.58–2.50
(1H, m, PCHAHB), 2.09–2.02 (2H, m, CH and PCHAHB), 1.53
(1H, br s, OH), 1.32–0.70 (3H, m, BH3) and 0.93 (3H, d, J 6.5,
Me); dC (125 MHz; CDCl3) 132.2 (d, J 9, PPh2 ortho), 132.0 (d, J
9, PPh2 ortho), 131.2 (d, J 2.5, PPh2 para), 131.1 (d, J 2.5, PPh2
para), 130.1 (d, J 55.5, PPh2 ipso), 130.0 (d, J 54.5, PPh2 ipso),
128.8 (d, J 10, PPh2 meta), 67.8 (d, J 9.0, CH2O), 31.8 (CH), 28.8
(d, J 36, PCH2) and 18.6 (d, J 5.5, Me); dP (162 MHz; CDCl3)
14.9–14.4 (m); m/z (ESI) 295 (58%, MNa+) and 281 (100, MNa -
BH3) (Found: MNa+, 295.14000. C16H22OPBNa requires M,
295.13990).
A representative sequence of reactions for the conversion of
compound 44 into compound 54 is given below:
(4S,2¢R)-3-[3¢-(Boronatodiphenylphosphinyl)-2¢-methylpropionyl]-
4-(phenylmethyl)oxazolidin-2-one 45
To a solution of hexamethyldisilazane (0.23 cm3, 1.1 mmol) in
dry THF (6 cm3), stirred at -78 ◦C under argon, was added
n-butyllithium (1.5 M solution in hexanes, 0.68 cm3, 1.05 mmol).
After 20 min, a solution of (S)-3-[3¢-(boronatodiphenylphos-
phinyl)propionyl]-4-(phenylmethyl)oxazolidin-2-one 44 (0.43 g,
1.0 mmol) in dry THF (6 cm3), at -78 ◦C under argon, was
added via cannula. After 1 h, methyl iodide (0.12 cm3, 2.0 mmol)
was added and the reaction mixture allowed to warm to room
temperature. After 42 h, the solvent was removed in vacuo and
the residue partitioned between dichloromethane (50 cm3) and
water (25 cm3). The organic layer was dried (Na2SO4), filtered
and the solvent removed in vacuo. The residue was purified by
flash column chromatography (SiO2, EtOAc–hexane 1 : 3, v/v)
to give oxazolidinone 45 (0.34 g, 76%) as an oil; [a]2D3 (c = 0.5,
CHCl3) +58.3; IR nmax (film)/cm-1 2926 (C–H), 2380 (B–H), 1776
(R)-[3-(Benzoyloxy)-2-methylpropyl]diphenylphosphine oxide 51
By the method of Pellon et al.,20 a solution of (R)-diphenyl(3-
hydroxy-2-methylpropyl)phosphine borane 48 (0.14 g, 0.50 mmol)
in toluene (1.5 cm3) was treated with DABCO (56 mg, 0.50 mmol)
◦
and the resulting mixture heated at 40 C for 18 h. The mixture
was treated with excess hydrogen peroxide solution and the residue
quenched with sodium metabisulfite. The mixture was partitioned
between dichloromethane (15 cm3) and water (15 cm3), the organic
layer dried (Na2SO4), filtered and the solvent removed in vacuo
to give crude (R)-diphenyl(3-hydroxy-2-methylpropyl)phosphine
oxide. dH (400 MHz; CDCl3) 7.79–7.69 (4H, m, Ph ortho), 7.56–
7.42 (6H, m, Ph), 3.61 (1H, dd, J 11 and 4, CHAHBO), 3.45 (1H, dd,
J 11.5 and 7.5, CHAHBO), 2.36 (1H, dd, J 14 and 8.5, PCHAHB),
2.32 (1H, dd, J 8.5 and 4, PCHAHB), 2.14–2.04 (1H, m, PCH2CH)
and 0.98 (3H, dd, J 7 and 1.5, Me); dC (125 MHz; CDCl3) 133.1
(d, J 99.5, Ph ipso), 131.9 (d, J 2.5, Ph para), 131.8 (d, J 98, Ph
ipso), 130.9 (d, J 9, Ph ortho), 130.6 (d, J 9.5, Ph ortho), 128.8 (d,
J 11.5, Ph meta), 128.7 (d, J 11.5, Ph meta), 68.3 (d, J 4.5, CH2O),
35.7 (d, J 69.5, PCH2), 32.1 (d, J 4, PCH2CH) and 19.9 (d, J
13, Me); dP (162 MHz; CDCl3) 35.4; m/z (ESI) 275 (55%, MH+)
(Found: MH+, 275.1190. C16H20O2P requires M, 275.1201). The
=
=
and 1694 (C O), 1605 (C C, Ph) and 1437 (P–Ph); dH (500 MHz;
CDCl3) 7.74–7.70 (2H, m, PPh2 ortho), 7.67–7.63 (2H, m, PPh2
ortho), 7.51–7.38 (6H, m, PPh2), 7.31–7.28 (2H, m, Ph), 7.26–7.23
(1H, m, Ph para), 7.15–7.13 (2H, m, Ph), 4.37 (1H, ddt, J 9.5, 7
and 3.5, CHN), 4.15–4.03 (3H, m, CH2O and CHMe), 3.17–3.10
(2H, m, CHAHBPh and PCHAHB), 2.74 (1H, dd, J 13.5 and 9.5,
PhCHAHB), 2.26 (1H, ddd, J 14.5, 12 and 2.5, PCHAHB), 1.30 (3H,
dd, J 7 and 1, Me) and 1.12–0.64 (3H, m, BH3); dC (125 MHz;
CDCl3) 175.2 (d, J 2.5, CONCO2), 152.7 (CONCO2), 135.1 (Ph
ipso), 132.5 (d, J 9, PPh2 ortho), 132.5 (d, J 9.5, PPh2 ortho), 131.4
(d, J 2.5, PPh2 para), 131.1 (d, J 2.5, PPh2 para), 129.4 (Ph), 129.3
(d, J 55.0, PPh2 ipso), 128.9 (d, J 10, PPh2 meta), 128.9 (Ph), 128.9
(d, J 55, PPh2 ipso), 128.6 (d, J 10, PPh2 meta), 127.3 (Ph para),
66.2 (CH2O), 55.3 (CHN), 37.8 (PhCH2), 33.3 (d, J 2, CHMe),
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1323–1328 | 1327
©

View Article Online
spectroscopic data are consistent with that reported previously.31
The crude product was dissolved in dichloromethane (4.5 cm3) and
triethylamine (0.10 cm3, 0.70 mmol), DMAP (30 mg, 0.25 mmol)
and benzoyl chloride (0.08 cm3, 0.70 mmol) were added. The
resulting solution was stirred at room temperature under argon
for 32 h. The mixture was washed with water (25 cm3) and
the aqueous layer extracted with ethyl acetate (4 ¥ 20 cm3).
The combined organic extracts were dried (Na2SO4), filtered and
the solvents removed in vacuo. The residue purified by flash
column chromatography (SiO2, CH2Cl2–MeOH 19 : 1, v/v) to
give phosphine oxide 51 (0.18 g, 96%) as an oil; [a]2D3.5 (c = 0.5,
CHCl3) +8.5; dH (500 MHz; CDCl3) 7.98–7.96 (2H, m, PhCO2
ortho), 7.78–7.73 (4H, m, PPh2 ortho), 7.55 (1H, t, J 7.5, Ph para),
7.51–7.39 (8H, m, Ph), 4.21–4.15 (2H, m, CH2O), 2.56 (1H, ddd,
J 15, 10.5 and 4.5, PCHAHB), 2.53–2.45 (1H, m, PCH2CH), 2.22
(1H, ddd, J 15, 12.5 and 8.5, PCHAHB) and 1.17 (3H, d, J 7, Me);
dC (125 MHz; CDCl3) 166.3 (CO2), 133.8 (d, J 98, PPh2 ipso),
133.0 (PhCO2 para), 132.8 (d, J 98, PPh2 ipso), 131.8 (d, J 3, PPh2
para), 131.7 (d, J 3, PPh2 para), 130.8 (d, J 9, PPh2 ortho), 130.6 (d,
J 9.5, PPh2 ortho), 130.1 (PhCO2 ipso), 129.5 (PhCO2), 128.7 (d,
J 11, PPh2 meta), 128.7 (d, J 12, PPh2 meta), 128.2 (PhCO2), 69.7
(d, J 12, CH2O), 33.2 (d, J 71.5, PCH2), 28.3 (d, J 3.5, PCH2CH)
and 18.6 (d, J 4.5, Me); dP (162 MHz; CDCl3) 31.6; m/z (ESI) 379
(100%, MH+) (Found: MH+, 379.1472. C23H24O3P requires M,
379.1463).
Acknowledgements
DSP thanks Torben & Alice Frimodts Fond, Brødrene Hartmanns
Fond and Direktør Ib Henriksens Fond for financial support.
We dedicate this paper to Professor A. B. Holmes in recognition
of Andy’s distinguished contributions to organic chemistry and
many years working together at Cambridge.
References
1 H. Lebel, J.-F. Marcoux, C. Molinaro and A. B. Charette, Chem. Rev.,
2003, 103, 977.
2 P. M. Ayrey and S. Warren, Tetrahedron Lett., 1989, 30, 4581.
3 T. Boesen, D. J. Fox, W. Galloway, D. S. Pedersen, C. R. Tyzack and S.
Warren, Org. Biomol. Chem., 2005, 3, 630.
4 L. Horner, H. Hoffmann and V. G. Toscano, Chem. Ber., 1962, 95,
536.
5 P. Wallace and S. Warren, J. Chem. Soc., Perkin Trans. 1, 1988,
2971.
6 R. A. Izydore and R. G. Ghirarde, J. Org. Chem., 1973, 38, 1790.
7 D. A. Evans, K. A. Woerpel and M. J. Scott, Angew. Chem., Int. Ed.
Engl., 1992, 31, 430.
8 W. S. Wadsworth Jr and W. D. Emmons, J. Am. Chem. Soc., 1961, 83,
1733.
9 E. E. Schweitzer and W. S. Creasy, J. Org. Chem., 1971, 36, 2379.
10 P. Wallace and S. Warren, Tetrahedron Lett., 1985, 26, 5713.
11 A. Nelson and S. Warren, J. Chem. Soc., Perkin Trans. 1, 1999,
3425.
12 A. Nelson and S. Warren, Tetrahedron Lett., 1996, 37, 1501.
13 D. J. Fox, S. Parris, D. S. Pedersen, C. R. Tyzack and S. Warren, Org.
Biomol. Chem., 2006, 4, 3108.
(1¢R,2¢R)-(2¢-Methylcyclopropyl)phenyl methanone 54
14 D. J. Fox, D. S. Pedersen and S. Warren, Org. Biomol. Chem., 2006, 4,
3113.
LDA was prepared by the addition of n-butyllithium (1.6 M
solution in hexanes, 1.25 cm3, 2.0 mmol) to a solution of
diisopropylam◦ine (0.28 cm3, 2.0 mmol) in dry THF (8.5 cm3),
stirred at -78 C under nitrogen. After 1 h, a solution of (R)-[3-
(benzoyloxy)-2-methylpropyl]diphenylphosphine oxide 51 (38 mg,
0.1 mmol) in dry THF (0.4 cm3), stirred at -78 ◦C under nitrogen,
was treated with the LDA solution (0.2 mol dm-3 solution in THF,
0.6 cm3, 0.12 mmol). After 1 h, the mixture was allowed to warm
to 0 ◦C and stirred for 48 h at this temperature. The reaction was
quenched with water (1 cm3) and the solvent removed in vacuo. The
residue was partitionedbetween water (2cm3) and ethyl acetate (3 ¥
5 cm3), the combined organic extracts dried (Na2SO4), filtered and
the solvents removed in vacuo. The residue was purified by flash
column chromatography (SiO2, CH2Cl2) to give cyclopropane 54
(10 mg, 60%) as an oil; [a]2D2 (c = 0.42, CHCl3) -77.1; Rf 0.5
15 D. A. Evans, Aldrichimica Acta, 1982, 15, 23.
16 J. R. Gage and D. A. Evans, Org. Synth., 1990, 68, 83.
17 D. R. Armstrong, M. G. Davidson, R. P. Davies, H. J. Mitchell, R. M.
Oakley, P. R. Raithby, R. Snaith and S. Warren, Angew. Chem., Int. Ed.
Engl., 1996, 35, 1942.
18 T. Hintermann and D. Seebach, Helv. Chim. Acta, 1998, 81,
2093.
19 S. D. Bull, S. G. Davies, S. Jones and H. J. Sanganee, J. Chem. Soc.,
Perkin Trans. 1, 1999, 387.
20 H. Brisset, Y. Gourdel, P. Pellon and M. Lecorre, Tetrahedron Lett.,
1993, 34, 4523.
21 T. Imamoto, T. Oshiki, T. Onozawa, T. Kusumoto and K. Sato, J. Am.
Chem. Soc., 1990, 112, 5244.
22 D. A. Evans, M. D. Ennis and D. J. Mathre, J. Am. Chem. Soc., 1982,
104, 1737.
23 E. N. Chauvel, C. Llorenscortes, P. Coric, S. Wilk, B. P. Roques and
M. C. Fourniezaluski, J. Med. Chem., 1994, 37, 2950.
24 A. V. Nemukhin, B. L. Grigorenko and A. A. Granovsky, Moscow
University Chemistry Bulletin, 2004, 45, 75.
25 R. Ditchfield, W. J. Hehre and J. A. Pople, J. Chem. Phys., 1971, 54,
724.
26 J. D. Dill and J. A. Pople, J. Chem. Phys., 1975, 62, 2921.
27 W. J. Hehre, R. Ditchfie and J. A. Pople, J. Chem. Phys., 1972, 56,
2257.
28 M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, M. S. Gordon,
D. J. Defrees and J. A. Pople, J. Chem. Phys., 1982, 77, 3654.
29 P. C. Harihara and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213.
30 P. M. W. Gill, Aust. J. Chem., 2001, 54, 661.
31 M. J. Doyle, D. Hall, P. R. Raithby, N. Skelton and S. Warren, J. Chem.
Soc., Perkin Trans. 1, 1993, 517.
(CH2Cl2); IR nmax (CH2Cl2)/cm-1 1664 (C O); dH (500 MHz;
=
CDCl3) 8.00–7.97 (2H, m, Ph ortho), 7.55 (1H, tt, J 7 and 1.5,
Ph para), 7.48–7.45 (2H, m, Ph meta), 2.39 (1H, dd, J 8 and 4,
CHCO), 1.64–1.56 (1H, m, CHMe), 1.48 (1H, ddd, J 8.5, 4.5 and
3.5, CHAHB), 1.21 (3H, d, J 6, Me) and 0.88 (1H, ddd, J 7.5, 6.5
and 3.5, CHAHB); dC (125 MHz; CDCl3) 200.1 (CO), 138.1 (Ph
ipso), 132.6 (Ph para), 128.4 and 127.9 (Ph), 26.4 and 21.3 (CH),
20.1 (CH2) and 18.3 (Me); m/z (EI) 160 (36%, M+) and 105 (100,
PhCO) (Found: M+, 160.08832. C11H12O requires M, 160.08882).
1328 | Org. Biomol. Chem., 2009, 7, 1323–1328
This journal is
The Royal Society of Chemistry 2009
©