Wittig rearrangement of lithiated allyl aryl ethers: A mechanistic study

Article abstract of DOI:10.1002/ejoc.200600304

At -75°C, α-lithiated allyl phenyl ether undergoes mainly the [1,2] Wittig rearrangement to afford, after acidic hydrolysis, 1-phenyl-2-propen-1-ol as the main product. A second metalation taking place at one of the ortho positions is the sole competing s

Full text of DOI:10.1002/ejoc.200600304

FULL PAPER
DOI: 10.1002/ejoc.200600304
Wittig Rearrangement of Lithiated Allyl Aryl Ethers: A Mechanistic Study
Sven Strunk[a] and Manfred Schlosser*[a,b]
Keywords: Allyl ethers / Intramolecular nucleophilic substitution / Metalation / Reaction mechanisms / Wittig
rearrangement
At –75 °C, α-lithiated allyl phenyl ether undergoes mainly the
[1,2] Wittig rearrangement to afford, after acidic hydrolysis,
1-phenyl-2-propen-1-ol as the main product. A second met-
alation taking place at one of the ortho positions is the sole
competing side reaction. Both, the significant decrease of the
isomerization rate upon the introduction of a tert-butyl sub-
stituent in the para position of the aromatic ring and the com-
plete absence of [1.4] rearrangement products suggest an in-
tramolecular addition/elimination process bringing about the
aryl migration. The first step, a nucleophilic attack of the α-
to the ipso-carbon atom generates a spiro-connected oxiranyl-
idene-cyclohexadienyllithium species. This short-lived inter-
mediate collapses to the final product, a lithium alkoxide, by
the nucleofugal departure of the oxygen atom which simulta-
neously binds the metal atom.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
Whereas the [3.2] mode offers the opportunity to ac-
complish stereocontrolled carbon–carbon linking reactions
and is therefore synthetically most attractive,[3–5] the [1.2]
mode was in the focus of numerous mechanistic studies.[1,6]
The principal features of the structural reorganization of α-
(alkoxy)alkyllithiums are nowadays widely accepted
(Scheme 2). During a two-stage process, the lithiated inter-
mediate splits into an alkyl radical and an O-lithiated alde-
hyde (or, rarely, ketone) anion-radical, a so-called “ketyl”,
before these fragments recombine to afford a lithium alk-
oxide.[1]
Introduction
What makes the Wittig rearrangement particularly fasci-
nating is its complexity. The manifold of mechanistic pos-
sibilities is multiplied when allyl entities are involved as
either the stationary or the migratory group.[1–6] In addition
to the ubiquitous [1,2] migration pattern, three more re-
arrangement modes can thus become operational by con-
necting the [1,4], [3,2] or [3,4] sites (as shown in the center
of Scheme 1).
Scheme 2. Homolytic cleavage of an α-(alkoxy)alkyllithium fol-
lowed by instantaneous radical-pair recombination.
This two-step mechanism satisfies all crucial experimen-
tal observables. The migratory aptitude of the dislocated
alkyl group mirrors the thermodynamical stability of the
transient radical obeying the order methyl Ͻ prim-alkyl Ͻ
sec-alkyl Ͻ tert-alkyl. As to be expected, 1-adamantyl be-
haves like tert-butyl,[7] whereas 1-(norbornyloxy)benzyllith-
ium[7] and 1-(apocamphyloxy)allyllithium[8] prove reluctant
to isomerize as the corresponding bridgehead radicals
would have to be pyramidal, hence would be energetically
unfavorable (Scheme 3).
The radical–ketyl pair must be extremely short-lived.
Only if the recombination occurs at approximately the same
rate as the radicals can tumble around one of their axes
within the confinements of a solvent cage, it becomes intelli-
gible why the configuration of a migrating sec-alkyl group[1]
Scheme 1. The four rearrangement modes available to metalated
allylic ethers (M standing for metal).
[a] Faculté des Sciences, Université de Lausanne,
1015 Lausanne, Switzerland
[b] Institute of Chemical Sciences and Engineering, Ecole Polytech-
nique Fédérale, BCh,
1015 Lausanne, Switzerland
Fax: +41-21-6939365
E-mail: manfred.schlosser@epfl.ch
Eur. J. Org. Chem. 2006, 4393–4397
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4393

S. Strunk, M. Schlosser
FULL PAPER
tives thereof[18] while stabilizing, if weakly, aryl radicals.[19]
All other common substituents favor both radical and carb-
anion formation although to a different degree. Thus, the
comparison between the migratory aptitudes of a phenyl
and a p-cymyl (4-tert-butylphenyl) group deemed us truly
conclusive.
Scheme 3. No migration of bridgehead-connected strained bicyclic
rings.
The metalation of both substrates, allyl phenyl ether and
allyl p-cymyl ether, proceeded quickly with sec-butyllithium
in tetrahydrofuran at –75 °C. After having trapped the lithi-
ated species with chlorotrimethylsilane, (Z)-trimethyl(3-
phenyloxy-2-propenyl)silane (1a, 77%) and (Z)-(3-p-cymyl-
oxy-2-propenyl)trimethylsilane (1b, 83%), were isolated.
When the reaction mixtures were brought to –25 °C, re-
arrangement set in. It turned out to be slow in the case of
the lithiated phenyl ether and even almost ten times slower
with the p-cymyl analog. Interception with chlorotrimethyl-
silane gave 1-phenyl-2-propenyl trimethylsilyl ether (3a,
25%) and 1-p-cymyl-2-propenyl trimethylsilyl ether (3b,
2.7%) as the main products. The bulk of the material was
obtained without isomerization in form of the silanes 1a
(58%) and 1b (70%) along with the bis(silanes) 2a (6.5%)
and 2b (9.4%) as by-products resulting from a twofold met-
alation at both the α-allyl and an ortho-aryl position
(Scheme 4). No other new compounds, in particular no (3-
aryl-2-propenyloxy)trimethylsilanes derived from [1,4] mi-
gration products 4, were identified.
is partially retained. The cyclopropylmethyl group, one of
the most rapidly ticking “molecular clocks”,[9] was found to
undergo very little, if any, ring opening when traveling from
the oxygen to the neighboring carbon atom.[8,10]
The rearrangement of α-aryloxy, as opposed to α-alk-
yloxy, substituted organolithium species has received much
less attention. G. Wittig et al.[11–13] had recognized in their
pioneering work a pronounced metal effect on the isomer-
ization rate. Lithium species react faster than sodium ana-
logs, and the latter faster than potassium derivatives.[11,12]
Furthermore, they demonstrated the intramolecularity of
the phenyl migration by a negative crossover experiment.[13]
Although the authors finally found a concerted O Ǟ C shift
of the phenyl ring more appealing, they took the alternative
intramolecular nucleophilic addition/elimination mecha-
nism explicitly into consideration.
This intramolecular nucleophilic displacement, already
tentatively favored by U. Schöllkopf,[1] was definitively res-
urrected when E. Grovenstein et al.[14] reinterpreted the re-
tention of the (Z)- and (E)-1-propenyl configurations ob-
served by V. Rautenstrauch, G. Büchi et al.[15] in the isomer-
ization of α-metalated benzyl ethers. The Swiss authors for
their part did not conceal their preference for another two-
stage scenario consisting of the heterolytic cleavage and
subsequent recombination of the fragments benzaldehydes
and propenyllithium. Finally, J. J. Eisch et al.[16] rejected
whatever mechanistic hypothesis so far proposed except the
homolytic dissociation/recombination sequence.
Results
We intended to shed new light on this unsolved issue by
applying two different probes. While monitoring to what
extent a para-alkyl substituent accelerates or retards the re-
arrangement, we wanted also to check whether or not at
least trace amounts of products attributable to the [1.4]
mode[17] were detectable.
The para substituent was selected to satisfy two criteria:
inertness toward strong bases and discrimination between
radicals and carbanions by stabilizing one and destabilizing
the other species. As we realized after long exploratory ex-
perimentation, tert-butyl is the sole substituent that meets
both conditions. Unlike methyl and primary or secondary
alkyl, it lacks benzylic C–H bonds and consequently resists
any attempt of metalation. At the same time, it neither un-
dergoes reductive degradation or elimination as halogen-
Scheme 4. Wittig rearrangement of α-lithiated phenyl and p-cymyl
2-propenyl ether.
Conclusions
The excellent mass balances of 96 and 91% with respect
bearing entities such as trifluoromethyl tend to do. More- to sec-butyllithium and 95 and 94% with respect to the allyl
over, it destabilizes carbanions and organometallic deriva- aryl ethers (taking into account recovered starting material)
4394
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4393–4397

Wittig Rearrangement of Lithiated Allyl Aryl Ethers: A Mechanistic Study
corroborate the claim that no 1,4-rearrangement products
FULL PAPER
The ultimate force driving the isomerization is the re-
were formed. They should have inevitably emerged from the placement of a weak[25,26] C–Li bond by a strong[24] O–Li
collapse of an aryl/ketyl radical pair. Thus, the homolysis/ bond. In order not to enter into high-energy regions, a so-
recombination mechanism can be ruled out. The substan- phisticated trajectory of metal delivery is mandatory. This
tial retardation of the isomerization rate caused by a para- will be the subject of forthcoming articles.[27,28]
alkyl substituent provides convincing evidence for a two-
step mechanism initiated by intramolecular nucleophilic ad-
Experimental Section
dition and terminated by nucleofugal elimination. It fea-
tures 8-vinyl-7-oxyspiro[5.2]octa-3,5-dien-2-yllithium com-
pounds 5 as the crucial intermediates in allyl aryl ether re-
arrangements (Scheme 5). This behavior is reminiscent of
the well-established mechanism of the Smiles isomeriza-
tion.[20–23]
1. Generalities: For working routine and abbreviations, see related
publications from this laboratory.[29–31] Unless specified otherwise,
the 1H NMR spectra were recorded at 360 MHz of samples dis-
solved in [D6]benzene or, if marked by an asterisk, in [D6]acetone.
2. Starting Materials and Compounds for Comparison
Allyl phenyl ether is commercial. Allyl p-cymyl ether and all refer-
ence compounds had to be prepared.
Trimethyl[(Z)-3-phenoxy-2-propenyl]silane (1a): At –75 °C, pre-
cooled tetrahydrofuran (25 mL) and allyl phenyl ether (2.7 g,
20 mmol) were added consecutively to sec-butyllithium (20 mmol)
from which the customary hydrocarbon solvent had been stripped
off beforehand. The mixture was kept 1 h at dry ice temperature
before being treated with chlorotrimethylsilane (2.5 mL, 2.2 g,
20 mmol). Upon distillation a colorless liquid was collected; b.p.
1
114–115 °C/10 Torr; n2D0 = 1.5062; yield 3.17 g (77%). H NMR: δ
= 7.1 (m, 2 H), 6.9 (m, 2 H), 6.83 (tt, J = 7.5, 1.0 Hz, 1 H), 6.28
(dt, J = 6.0, 1.2 Hz, 1 H), 4.71 (td, J = 8.7, 6.0 Hz, 1 H), 1.65 (dd,
J = 8.6, 1.2 Hz, 2 H), 0.04 (s, 9 H) ppm. MS: m/z (%) = 206 (4)
[M+], 191 (4), 73 (100). C12H18OSi (206.36): calcd. C 69.84, H 8.79;
found C 70.15, H 8.91.
Trimethyl[(Z)-2-(3-trimethylsilyl-1-propenyl)oxyphenyl]silane (2a):
The same protocol as described in the preceding paragraph was
applied, but 2 equiv. (40 mmol) of sec-butyllithium were allowed to
react with the substrate for 20 h in a dry ice/methanol bath before
Scheme 5. Rearrangement of α-metalated allyl aryl ethers accord-
ing to the intramolecular nucleophilic addition/elimination (rather
than the homolytic cleavage/recombination) mechanism.
the mixture was treated with chlorotrimethylsilane (4.9 mL, 4.3 g,
20
40 mmol); colorless oil; b.p. 84–85 °C/1 Torr; n
= 1.5018; yield
D
3.12 g (75%). 1H NMR*: δ = 7.4 (m, 2 H), 7.05 (td, J = 7.4, 1.0 Hz,
1 H), 6.99 (d, J = 8.2 Hz, 1 H), 6.43 (dt, J = 8.8, 6.0 Hz, 1 H), 4.92
(td, J = 8.8, 6.0 Hz, 1 H), 1.66 (dd, J 8.8, 1.2 Hz, 2 H), 0.32 (s, 9
H), 0.05 (s, 9 H) ppm. MS: m/z (%) = 206 (1) [M+], 175 (25), 73
(100). C15H26OSi (278.54): calcd. C 64.68, H 9.41; found C 64.79,
H 9.41.
Although the two competing intermediates, the hypo-
thetical ketyl/aryl radical pair and the oxaspirooctadienyl-
lithium 5 are not the rate-determining transition states, they
should lie at almost the same energetic level. Relying on
this plausible assumption we can compare the activation
required for either process. The homolytic scission of the
aryl–oxygen linkage costs some 95 kcal/mol, that of the al-
lyl–lithium interaction about 50 kcal/mol. After deduction
Trimethyl[(1-phenyl-2-propenyl)oxy]silane (3a): The mixture of 1-
phenyl-2-propen-1-ol[32] (5.0 g, 37 mmol), chlorotrimethylsilane
(6.2 mL, 5.4 g, 50 mmol) and triethylamine (14 mL, 10 g, 0.10 mol)
was kept 12 h at +25 °C before being diluted with hexanes (20 mL)
of the enthalpy of the newly formed O–Li bond of 120 kcal/ and washed with ice-cold 5% hydrochloric acid (2×10 mL) and a
mol[24] and the improved resonance due to the conversion
of the allyl into the ketyl species by 5 kcal/mol remains a
deficit of 20 kcal/mol. This barrier shrinks to approximately
15 kcal/mol if the intramolecular nucleophilic substitution
saturated aqueous solution (2×20 mL) of sodium hydrogen car-
bonate. Distillation gave a colorless oil; b.p. 52–56 °C/0.5 Torr;
n2D0 = 1.4832; yield 6.11 g (80%). H NMR: δ = 7.4 (m, 2 H), 7.2
1
(m, 2 H), 7.07 (tt, J = 7.0, 1.0 Hz, 1 H), 5.94 (ddd, J = 16.8, 10.2,
6.0 Hz, 1 H), 5.26 (ddd, J = 16.8, 2.0, 1.5 Hz, 1 H), 5.11 (broad d,
mechanism is elicited. At the stage of intermediate 5 no
J = 6.0 Hz, 1 H), 4.96 (ddd, J = 10.2, 2.0, 1.5 Hz, 1 H), 0.15 (s, 9
bond scission nor lithium transfer has yet occurred. The
spiro annulation builts up ring strain in the order of 25 kcal/
mol. It is only partially compensated by the increase of the
resonance energy in pentadienyllithium as opposed to allyl-
lithium (30 vs. 20 kcal/mol). Even if the estimates made
above are admittedly crude, they support the experimental
findings which suggest an intramolecular nucleophilic sub-
stitution mechanisms to be operative and rule out a homo-
lytic dissociation/recombination process.
H) ppm. MS: m/z (%) = 206 (9) [M+], 191 (8), 117 (38), 73 (100).
C12H18OSi (206.36): calcd. C 69.84, H 8.79; found C 70.03, H 8.75.
Trimethyl[(Z)-(1-phenyl-1-propenyl)oxy]silane: Precooled (–75 °C)
tetrahydrofuran (10 mL) and 1-phenyl-2-propen-1-ol[32] (0.67 g,
5.0 mmol) were added consecutively to sec-butyllithium (5.5 mmol)
from which before the commercial hydrocarbon solvent had been
stripped off. After 170 h (1 week) at +25 °C, the mixture was
poured into hexanes (20 mL) and washed with 2% sulfuric acid
(2×5 mL), a saturated aqueous solution (2×5 mL) of sodium hy-
Eur. J. Org. Chem. 2006, 4393–4397
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4395

S. Strunk, M. Schlosser
FULL PAPER
drogen carbonate and brine (5 mL). Upon evaporation of the vola-
tiles and distillation of the residue, trimethyl[(1-phenyl-1-propenyl)-
oxy]silane and its isomer 3a were collected in a 9:1 ratio; b.p. 50–
56 °C/0.5 Torr; yield 0.86 g (83%). The main component was puri-
fied by preparative gas chromatography (3 m, 10% Apiezon-L,
200 °C); n2D0 = 1.5040. 1H NMR: δ = 7.5 (m, 2 H), 7.2 (m, 2 H),
7.06 (tt, J = 7.2, 1.2 Hz, 1 H), 5.28 (q, J = 6.8 Hz, 1 H) 1.73 (d, J
= 6.8 Hz, 3 H), 0.09 (s, 9 H). MS: m/z (%) = 206 (12) [M+], 205
(18), 105 (21), 75 (100). C12H18OSi (206.36): calcd. C 69.84, H 8.79;
found C 69.85, H 8.75.
3. Rearrangement of Allyl Phenyl Ether and Allyl p-Cymyl Ether
At –75 °C, precooled tetrahydrofuran (9.0 mL) and allyl phenyl
ether (0.67 g, 5.0 mmol) were added consecutively to sec-butyllith-
ium (5.0 mmol) from which the commercial hydrocarbon solvent
had beforehand been stripped off. The mixture was kept for 2 h
at –25 °C before being treated with chlorotrimethylsilane (1.3 mL,
1.1 g, 10 mmol) and N,N,NЈ,NЈ-tetramethylethylenediamine
(1.5 mL, 1.2 g, 10 mmol). Gas chromatography (2 m, 5% silicon
rubber SE-30, 80 °C [5 min] Ǟ 200 °C [10 °C/min]; 2 m, 5% hydro-
carbon Apiezon-L, same temperature program; undecane as the
calibrated internal standard) revealed the presence of the silane 1a
(58%), the bis(silane) 2a (6.5%), the rearranged trimethylsilyl ether
3a (19%) and its isomerization product trimethyl[(1-phenyl-1-pro-
penyl)oxy]silane (5.8%). Not even trace amounts of trimethyl[(Z)-
(3-phenyl-1-propenyl)oxy]silane were detected. Some unconsumed
allyl phenyl ether (5.1%) was identified too.
Trimethyl[(Z)-(3-phenyl-1-propenyl)oxy]silane:
A solution of 3-
phenylpropanal[33] (13 g, 0.10 mol), chlorotrimethylsilane (15 mL,
13 g, 0.12 mol) and triethylamine (35 mL, 31 g, 0.25 mol) in di-
methylformamide (50 mL) was heated to reflux for 24 h. After the
addition of hexanes (0.10 L), the mixture was washed with ice-cold
hydrochloric acid (2 ×0.10 L), a saturated aqueous solution
(2×0.10 L) of sodium hydrogen carbonate and brine (0.10 L). Dis-
tillation afforded the product as a 2:1 mixture of (Z) and (E) iso-
mers; b.p. 63–72 °C/1 Torr; yield 14.2 g (69%). The main compo-
nent was purified by preparative gas chromatography (3 m, 10%
The same protocol was applied to allyl p-cymyl ether (0.96 g,
5.0 mmol). Gas chromatographic analysis (same columns and tem-
perature profiles as above; tetradecane as the calibrated internal
standard) identified the silane 1b (70%), the bis(silane) 2b (9.4%)
and the rearranged trimethylsilyl ether 3b (2.7%). A moderate
quantity (12%) of allyl p-cymyl ether was recovered.
1
Apiezon-L, 200 °C). H NMR: δ = 7.2 (m, 4 H), 7.05 (tt, J = 7.0,
1.2 Hz, 1 H), 6.19 (dt, J = 5.8, 1.5 Hz, 1 H), 4.75 (td, J = 7.5,
5.8 Hz, 1 H), 3.59 (dd, J = 7.5, 1.2 Hz, 2 H), 0.05 (s, 9 H). MS:
m/z (%) = 206 (11) [M+], 191 (9), 117 (20), 73 (100). C12H18OSi
(206.36): calcd. C 69.84, H 8.79; found C 70.27, H 8.84.
Acknowledgments
Allyl p-Cymyl Ether (Allyl 4-tert-Butylphenyl Ether): Potassium
carbonate (28 g, 0.20 mol) was added to a solution of 4-tert-bu-
tylphenol (30 g, 0.20 mol) and allyl bromide (24 g, 0.20 mol) in an-
hydrous acetone (25 mL). The mixture was heated to reflux for 9 h
before being washed with water (0.10 L) and a 2.0  aqueous solu-
tion (3×25 mL) of sodium hydroxide. The product was isolated by
distillation under reduced pressure; b.p. 68–69 °C/0.5 Torr (lit.[34]
115–116 °C/3 Torr); 1.5069 (lit.[34] 1.5065); yield 20.6 g (54%). 1H
NMR: δ = 7.3 (m, 2 H), 6.9 (m, 2 H), 6.06 (ddt, J = 17.5, 10.5,
5.5 Hz, 1 H), 5.42 (dq, J = 17.5, 1.5 Hz, 1 H), 5.28 (dq, J = 10.5,
1.5 Hz, 1 H), 4.52 (dt, J = 5.5, 1.5 Hz, 2 H), 1.29 (s, 9 H) ppm.
MS: m/z (%) = 190 (35) [M+], 175 (100). C13H18O (190.29): calcd.
C 82.06, H 9.54; found C 82.34, H 9.47.
This work was supported by the Swiss National Science Founda-
tion, Bern (grant 20- 100Ј336-02).
[1] U. Schöllkopf, Angew. Chem. 1970, 82, 795–805; Angew. Chem.
Int. Ed. Engl. 1970, 9, 763–773.
[2] T. Nakai, K. Mikami, Chem. Rev. 1986, 86, 885–902.
[3] K. Mikami, T. Nakai, Synthesis 1991, 594–604.
[4] T. Nakai, K. Mikami, Org. React. 1994, 46, 105–209.
[5] T. Nakai, K. Tomooka, Pure Appl. Chem. 1997, 69, 595–600.
[6] K. Tomooka, H. Yamamoto, T. Nakai, Liebigs Ann./Recueil
1997, 1275–1281.
[7] P. T. Lansbury, V. A. Pattison, J. D. Sidler, J. B. Bieber, J. Am.
Chem. Soc. 1966, 88, 78–84.
[8] S. Strunk, Doctoral Dissertation, Université de Lausanne, 1987,
(Z)-(3-p-Cymyloxy-2-propenyl)trimethylsilane (1b): Prepared, anal-
ogously as the phenoxy compound 1a, from allyl p-cymyl ether
(3.8 g, 20 mmol); b.p. 101–102 °C/0.5 Torr; n2D0 = 1.5022; yield
pp. 49–50.
[9] D. Maillard, D. Forrest, K. U. Ingold, J. Am. Chem. Soc. 1976,
98, 7024–7026.
1
4.34 g (83%). H NMR: δ = 7.2 (m, 2 H), 7.0 (m, 2 H), 6.32 (dt, J
[10] P. T. Lansbury, V. A. Pattison, J. Am. Chem. Soc. 1962, 84,
4295–4298.
= 6.0, 1.2 Hz, 1 H), 4.76 (dt, J = 8.7, 6.2 Hz, 1 H), 1.71 (dd, J =
8.6, 1.2 Hz, 2 H), 1.21 (s, 9 H), 0.06 (s, 9 H) ppm. MS: m/z (%) =
262 (8) [M+], 247 (17), 205 (45), 73 (100). C16H26OSi (262.47):
calcd. C 73.22, H 9.99; found C 73.31, H 9.93.
[11] G. Wittig, W. Happe, Justus Liebigs Ann. Chem. 1947, 557,
205–220.
[12] G. Wittig, R. Clausnizer, Justus Liebigs Ann. Chem. 1954, 588,
145–166.
{5-tert-Butyl-2[(Z)-(3-trimethylsilyl-1-propenyl)oxy]}trimethylsilane
(2b): Prepared and isolated as described for the analog 2a; b.p. 123–
127 °C/1 Torr; n2D0 = 1.4996; yield 6.23 g (93%). 1H NMR*: δ =
7.48 (d, J = 2.5 Hz, 1 H), 7.43 (dd, J = 8.5, 2.6 Hz, 1 H), 6.91 (d,
J = 8.5 Hz, 1 H), 6.41 (dd, J = 6.0, 1.2 Hz, 1 H), 4.87 (td, J = 8.7,
6.2 Hz, 1 H), 1.66 (dd, J = 8.7, 1.2 Hz, 2 H), 1.31 (s, 9 H), 0.32 (s,
9 H), 0.05 (s, 9 H). C19H34OSi2 (334.65): calcd. C 68.19, H 10.24;
found C 68.18, H 10.28.
[13] G. Wittig, E. Stahnecker, Justus Liebigs Ann. Chem. 1957, 605,
69–93.
[14] E. Grovenstein, K. W. Black, S. G. Goel, R. L. Hughes, J. H.
Northrop, D. L. Streeter, D. VanDerveer, J. Org. Chem. 1989,
54, 1671–1679.
[15] V. Rautenstrauch, G. Büchi, H. Wüest, J. Am. Chem. Soc. 1974,
96, 2576–2580.
[16] J. J. Eisch, C. A. Kovacs, S.-G. Rhee, J. Organomet. Chem.
1974, 65, 289–301.
[17] H. Felkin, A. Tambuté, Tetrahedron Lett. 1969, 10, 821–822.
[18] M. Schlosser, O. Desponds, R. Lehmann, E. Moret, G.
Rauchschwalbe, Tetrahedron 1993, 49, 10175–10203.
[19] C. G. Swain, W. H. Stockmayer, J. T. Clarke, J. Am. Chem. Soc.
1950, 72, 5426–5433.
[(1-p-Cymyl-1-propenyl)oxy]trimethylsilane (3b): Prepared and iso-
lated as described for the phenyl analog 3a; b.p. 82–84 °C/0.5 Torr;
1
n2D0 = 1.4841; yield 7.09 g (73%). H NMR: δ = 7.4 (m, 2 H), 7.3
(m, 2 H), 6.02 (ddd, J = 17.0, 10.5, 6.0 Hz, 1 H), 5.33 (ddd, J =
10.5, 2.0, 1.5 Hz, 1 H), 5.18 (broad d, J = 6.0 Hz, 1 H), 5.02 (ddd,
J = 10.5, 2.0, 1.5 Hz, 1 H), 1.21 (s, 9 H), 0.11 (s, 9 H). MS: m/z
(%) = 262 (5) [M+], 247 (19), 205 (100), 73(55). C16H26OSi (262.47):
calcd. C 73.22, H 9.99; found C 73.51, H 9.97.
[20] J. Sauer, R. Huisgen, Angew. Chem. 1960, 72, 294–315, specifi-
cally p. 314.
[21] I. G. C. Coutts, M. R. Southcott, J. Chem. Soc., Perkin Trans.
1 1990, 767–771.
4396
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4393–4397

Wittig Rearrangement of Lithiated Allyl Aryl Ethers: A Mechanistic Study
FULL PAPER
[22] K. Bowden, P. R. Williams, J. Chem. Soc., Perkin Trans. 2 1991, [29] C. Bobbio, M. Schlosser, Eur. J. Org. Chem. 2001, 4533–4536.
215–224.
[30] M. Masson, M. Schlosser, Eur. J. Org. Chem. 2005, 4401–4405.
[31] F. Leroux, O. Lefebvre, M. Schlosser, Eur. J. Org. Chem. 2006,
in press.
[32] H. E. Ramsden, J. R. Leebrick, S. D. Rosenberg, E. H. Miller,
J. J. Walburn, A. E. Balint, R. Cserr, J. Org. Chem. 1957, 22,
1602–1605.
[23] V. J. Huber, R. A. Bartsch, Tetrahedron 1998, 54, 9281–9288.
[24] J. B. Pedley, E. M. Marshall, J. Phys. Chem. Ref. Data 1983,
12, 967–1031 [Chem. Abstr. 1984, 100, 145949].
[25] M. Schlosser, in: Organometallics in Synthesis: A Manual (Ed.:
M. Schlosser), 2nd edition, Wiley, Chichester, 2002, pp. 1–352,
specifically pp. 46–49.
[26] J. W. Ochterski, G. A. Petersson, K. B. Wiberg, J. Am. Chem.
Soc. 1995, 117, 11299–11308, specifically p. 11307.
[27] M. Schlosser, F. Bailly, manuscript in preparation.
[28] F. Bailly, R. Scopelliti, M. Schlosser, manuscript in prepara-
tion.
[33] M. Delépine, C. Hanegraaff, Bull. Soc. Chim. Fr. 1937, 5, 2087–
2093.
[34] A. B. Sen, R. P. Rastogi, J. Indian Chem. Soc. 1953, 30, 355–
358 [Chem. Abstr. 1954, 48, 10649e].
Received: April 7, 2006
Published Online: August 2, 2006
Eur. J. Org. Chem. 2006, 4393–4397
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4397