Part I: The development of the catalytic wittig reaction

Article abstract of DOI:10.1002/chem.201301444

We have developed the first catalytic (in phosphane) Wittig reaction (CWR). The utilization of an organosilane was pivotal for success as it allowed for the chemoselective reduction of a phosphane oxide. Protocol optimization evaluated the phosphane oxide precatalyst structure, loading, organosilane, temperature, solvent, and base. These studies demonstrated that to maintain viable catalytic performance it was necessary to employ cyclic phosphane oxide precatalysts of type 1. Initial substrate studies utilized sodium carbonate as a base, and further experimentation identified N,N-diisopropylethylamine (DIPEA) as a soluble alternative. The use of DIPEA improved the ease of use, broadened the substrate scope, and decreased the precatalyst loading. The optimized protocols were compatible with alkyl, aryl, and heterocyclic (furyl, indolyl, pyridyl, pyrrolyl, and thienyl) aldehydes to produce both di- and trisubstituted olefins in moderate-to-high yields (60-96 %) by using a precatalyst loading of 4-10 mol %. Kinetic E/Z selectivity was generally 66:34; complete E selectivity for disubstituted α,β-unsaturated products was achieved through a phosphane-mediated isomerization event. The CWR was applied to the synthesis of 54, a known precursor to the anti-Alzheimer drug donepezil hydrochloride, on a multigram scale (12.2 g, 74 % yield). In addition, to our knowledge, the described CWR is the only transition-/heavy-metal-free catalytic olefination process, excluding proton-catalyzed elimination reactions. A point of difference: By utilizing an organosilane to chemoselectively reduce a phosphane oxide precatalyst to a phosphane (see scheme), the first catalytic (in phosphane) Wittig reaction has been developed. The methodology has been applied to the synthesis of 22 disubstituted and 24 trisubstituted olefins, including a multigram synthesis of a precursor to the anti-Alzheimer drug donepezil hydrochloride.

Full text of DOI:10.1002/chem.201301444

FULL PAPER
DOI: 10.1002/chem.201301444
Part I: The Development of the Catalytic Wittig Reaction
Christopher J. OꢀBrien,*[a] Zachary S. Nixon,[b] Andrew J. Holohan,[a]
Stephen R. Kunkel,[a] Jennifer L. Tellez,[c] Bryan J. Doonan,[a] Emma E. Coyle,[a]
Florie Lavigne,[a] Lauren J. Kang,[c] and Katherine C. Przeworski[c]
Dedicated to Georg Wittig (1897–1987) Nobel Prize 1979
Abstract: We have developed the first
catalytic (in phosphane) Wittig reaction
(CWR). The utilization of an organosi-
lane was pivotal for success as it al-
lowed for the chemoselective reduction
of a phosphane oxide. Protocol optimi-
zation evaluated the phosphane oxide
precatalyst structure, loading, organosi-
lane, temperature, solvent, and base.
These studies demonstrated that to
maintain viable catalytic performance
it was necessary to employ cyclic phos-
phane oxide precatalysts of type 1. Ini-
tial substrate studies utilized sodium
carbonate as a base, and further experi-
mentation identified N,N-diisopropyl-
G
10 mol%. Kinetic E/Z selectivity was
generally 66:34; complete E selectivity
for disubstituted a,b-unsaturated prod-
ucts was achieved through a phos-
phane-mediated isomerization event.
The CWR was applied to the synthesis
of 54, a known precursor to the anti-
Alzheimer drug donepezil hydrochlor-
ide, on a multigram scale (12.2 g, 74%
yield). In addition, to our knowledge,
the described CWR is the only transi-
tion-/heavy-metal-free catalytic olefina-
tion process, excluding proton-cata-
lyzed elimination reactions.
Introduction
tility of olefins, a vast amount of effort has been directed at
their construction. Arguably, other than direct elimination,[2]
there are currently four general methodologies for the rou-
tine and reliable formation of alkenes:[3] 1) Wittig,[4] 2) Pe-
terson,[5] 3) Julia–Lythgoe[6]/Julia–Kocienski[7] olefination re-
actions, and 4) metathesis.[8] In addition to these reactions,
the Heck reaction may be utilized to produce certain alkene
structures.[9] Of the olefinations listed, only metathesis and
the Heck reaction are catalytic, both require alkene starting
materials and use a transition-metal catalyst. Strictly, meta-
thesis and the Heck reaction can be considered as alkene
augmentation processes rather than olefinations. One stoi-
chiometric protocol that offers the opportunity to evolve
into a catalytic process is the Wittig reaction.[4] First dis-
closed in a 1953 report by Wittig and Geissler,[4a] this reac-
tion involves the treatment of an aldehyde or ketone with a
phosphonium ylide, thus resulting in an alkene concurrent
with a phosphane oxide byproduct.[4] Since Wittigꢀs discov-
ery the reliability of his reaction has resulted in its extensive
use in organic chemistry.[10] Nonetheless, the Wittig reaction
suffers from several impediments, namely that the process is
stoichiometric, which often results in difficulty in the remov-
al of the phosphane oxide byproduct. Moreover, lack of a
catalytic Wittig protocol, due to cost/benefit, removes the
possibility to control the olefination event by alteration of
phosphane structure. This deficiency is unfortunate because
The carbon–carbon double bond presents a myriad of op-
portunities for the synthetic chemist.[1] Olefins, other than
offering a degree of structural rigidity, are the basis for
many robust synthetic methodologies.[1] Therefore, an
alkene precursor in a synthetic route may act as a pivot to
many distinct structural classes or offer an alternative when
problems present. Unsurprisingly, due to the synthetic versa-
[a] Dr. C. J. OꢀBrien, A. J. Holohan, S. R. Kunkel, B. J. Doonan,
Dr. E. E. Coyle, Dr. F. Lavigne
School of Chemical Sciences
Dublin City University
Glasnevin, Dublin 9 (Ireland)
E-mail: christopher.obrien@dcu.ie
Homepage: http://webpages.dcu.ie/~obrienc/OBrien_Group/
Home.html
[b] Dr. Z. S. Nixon
Sigma-Aldrich Chemical Corporation
Catalysis Development Group
6000 North Teutonia Avenue, Milwaukee, WI 53209 (USA)
[c] J. L. Tellez, L. J. Kang, K. C. Przeworski
Department of Chemistry and Biochemistry
The University of Texas at Arlington
Box 19065, Arlington, TX 76019 (USA)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201301444.
Chem. Eur. J. 2013, 19, 15281 – 15289
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
15281

the phosphane structure has been demonstrated to have a
substantial impact on the stereochemical outcome of the re-
action.[11] Thus, carefully designed phosphanes may yield se-
lective processes, and these endeavors would be enhanced if
the phosphane was employed catalytically.[11e]
The obstacles to the development of a catalytic Wittig re-
action are formidable, and the construction of a catalytic
cycle relies on the completion of four discrete steps
(Scheme 1): 1) formation of the phosphonium ylide precur-
Results and Discussion
Fundamentally, the challenge in developing a CWR is the
chemoselective reduction of the P=O bond in the presence
of other reactive functionalities, in this case an aldehyde or
ketone and an alkene. At the time, a literature review re-
vealed there were few protocols for phosphane oxide reduc-
tion.[24–26] Of these methods, most would be unsuitable as a
reductant in a proposed CWR: lithium aluminum hydride[25]
is a harsh relatively unselective reducing agent and tri-
silane (with or without an amine base)[24] has selectivi-
ACHUTNGRENcNUG hloroACHTUTGNRENNUGN
ꢀ
ty concerns that stem from the lability of the Si Cl bond
and control of the phosphorus center during the reduction.
Yet the solution may be found in a structural relative of tri-
chlorosilane; that is, both phenylsilane and diphenylsilane
are known to reduce phosphane oxides and are likely to be
compatible with a catalytic Wittig process (Scheme 1). Fur-
thermore, these organosilanes have been shown to reduce
phosphane oxides with retention of configuration at the
phosphorus center.[27] The preservation of the stereochemi-
cal integrity at the phosphorus center during reduction is
important if the structure of the phosphane is to be used
later to impart diastereoselectivity on the Wittig reaction.
Additionally, aryl silanes are unlikely to reduce any alde-
hyde or ketone because hydrosilylation normally requires a
transition metal.[28]
Following the identification of a suitable class of reducing
agent, the next consideration was the nature of the phos-
phane oxide precatalyst. One may consider triphenylphos-
phane oxide, itself a byproduct of many stoichiometric
Wittig reactions, to be ideal. However, the ease of the re-
duction of the phosphane oxide was a crucial factor. From
the outset, a key design criterion for the CWR was that it
would be complete in 24 hours. To fulfill this condition, if a
precatalyst loading of 10 mol% was employed then the cata-
lyst must cycle once every 2.4 hours. We felt that use of tri-
phenylphosphane oxide would not satisfy this requirement.
The employment of cyclic phosphane oxides was more
promising as ring strain is known to significantly aid the
ease of reduction.[29] Following this rationale, commercially
available 3-methyl-1-phenylphospholene-1-oxide (1a) was
considered, but the presence of the double bond could
prove troublesome; therefore, this double bond was re-
moved by hydrogenation, thus yielding 3-methyl-1-phenyl-
phospholane-1-oxide (1b) as a mixture of diastereomers
(Scheme 2).
Scheme 1. Proposed catalytic Wittig reaction.
sor, typically a phosphonium salt;[10,11] 2) generation of the
phosphonium ylide, normally by deprotonation;[10–12] 3) olefi-
nation with the concomitant generation of a phosphane
oxide;[10–12] and 4) reduction of the phosphane oxide byprod-
uct to produce a phosphane to re-enter the cycle. Most chal-
lenging of the above processes is the essential chemoselec-
tive reduction of the phosphane oxide, which must be ac-
complished in the presence of either aldehyde or ketone
starting materials and an alkene product. Amelioration of
this chemoselective reduction problem could be achieved by
the exchange of the phosphorus atom with arsenic,[13] telluri-
um,[14] or antimony[15] because their oxides, due to bond
strength, are appreciably easier to reduce.[16] Indeed, this
strategy led to catalytic Wittig-type processes that employ
arsines and tellurides.[17] A significant drawback to the
broad adoption of the aforementioned methodologies are
the high toxicity and carcinogenicity of arsenic,[18] telluri-
um,[19] and antimony compounds.[20] Environmental anthro-
pogenic contamination, particularly of groundwater, is a
major concern if these reactions were performed on a large
scale.[21] The catalytic use of a phosphane would not suffer
from these limitations; hence, a Wittig reaction catalytic in
phosphane would find wider employment. Furthermore,
such a process would marry the power of the Wittig olefina-
tion reaction to the synthetic benefits of a catalytic protocol
without the poisoning issues that can plague transition-
metal-catalyzed processes.[22] Recently, our laboratory dis-
closed the first catalytic Wittig reaction (CWR),[23] and
herein we present in full our studies that led to the develop-
ment of this protocol.
Scheme 2. Preparation of 1b through hydrogenation (see the Supporting
Information for full details).
15282
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 15281 – 15289

Organocatalysis of the Wittig Reaction
FULL PAPER
Figure 1. Optimization of organosilane reduction of phosphane oxide 1b.
Table 1. Base study using phosphonium salt.[a]
Subsequently, the reduction of 1b was evaluated by using
diphenylsilane, phenylsilane, and trimethoxysilane in
[D8]toluene at 60, 80, and 1008C (Figure 1). From these ex-
periments, it was clear that all three silane compounds were
viable reducing agents at 1008C because the conversion into
a phosphane compound was nearly complete in just 20 mi-
nutes. At lower temperatures (i.e., 60 and 808C), the reduc-
tion was noticeably slower, with trimethoxysilane yielding
the highest conversions. These results demonstrated that a
reaction temperature of 1008C was necessary for a viable
CWR. In addition, it was confirmed that reduction proceed-
ed with no inversion at the phosphorus center (see the Sup-
porting Information).
After the identification of a suitable reduction strategy,
the next consideration was the actual Wittig reaction, reliant
on ylide formation through deprotonation of the phosphoni-
um salt. For this step, the choice of base is critical. The base
must have a pKa value (of the conjugate acid) that is suffi-
cient to deprotonate the ylide-forming proton of the phos-
phonium salt, yet not lead to incompatibilities with other re-
agents (such as the silane, organohalide, or olefin). Phospho-
nium salt 3, formed from the reaction of phosphane 2 with
methyl bromoacetate, was prepared and evaluated with a
series of bases. It was anticipated that the ylide-forming
proton of this salt would have a pKa value of approximately
9, therefore the use of common bases of pKa 9–12 (for con-
jugate acid) were investigated (Table 1). These studies iden-
tified sodium carbonate as the base of choice by combining
high yield and ease of handling.[30]
Entry
Base
Conversion [%][b]
E/Z[c]
1
2
3
4
K3PO4
100
75:25
75:25
75:25
62:38
K2CO3
Na2CO3
Na2SO3
95 (90)
98 (90)
54
[a] Benzaldehyde (1.0 mmol), 3 (1.1 mmol), base (1.5 mmol), and toluene
(3.0m). [b] Conversions were determined by using H NMR spectroscopy;
selected examples were purified to determine the yield of the isolated
product (shown in parentheses). [c] E/Z ratio was determined by
1H NMR spectroscopic analysis of the unpurified reaction mixture.
1
Table 2. Optimization of organosilane for CWR.[a]
Entry
Silane
Yield [%][b]
E/Z[c]
n.d.
>95:5
>95:5
70:30
1
2
3
4
Ph3SiH
Ph2SiH2
PhSiH3
trace
75
46
(MeO)3SiH
61
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.3 mmol), 1b
(0.1 mmol), sodium carbonate (1.5 mmol), silane (1.1 mmol), and toluene
(3.0m). [b] Yields were determined by GCMS/MS against a calibrated in-
ternal standard (undecane) and were performed in duplicate. [c] E/Z
ratio was determined by GCMS/MS. n.d.=not determined.
Next, the full CWR was assembled by combining the opti-
mized conditions for catalyst, temperature, reductant, and
base (Table 2). Interestingly, although trimethoxysilane gave
the best results in the reduction studies, this result did not
translate into the full reaction system, and diphenylsilane
give the optimal yield and selectivity. After having demon-
strated the first catalytic in phosphane Wittig reaction, we
sought to optimize the reaction further by alteration of the
solvent (Table 3) and reaction temperature (Table 4). Aceto-
nitrile and toluene were shown to give the greatest yields
after both 5 and 24 hours, respectively. However, the use of
acetonitrile at a temperature greater than 808C necessitated
the use of a sealed vessel, thus toluene offered greater ease
of use. Increasing the reaction temperature from 80 to
1008C did not lead to a significant variation in yield, but
gave enhanced diastereoselectivity (Table 4, entries 3 and 4).
It was observed that the E/Z selectivity of the reaction after
just 5 hours was generally 66:34, whereas the reaction was
E-selective after 24 hours (cf. Table 3). This finding suggests
Chem. Eur. J. 2013, 19, 15281 – 15289
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15283

C. J. OꢀBrien et al.
Table 3. Optimization of solvent.[a]
Table 5. Catalyst loading study.[a]
Entry
Solvent
t [h]
Yield [%][b]
E/Z[c]
66:34
Entry
Loading
1b [mol%]
Yield [%][b]
E/Z[c]
1
2
3
4
5
6
7
8
toluene
toluene
5
24
5
24
5
24
5
24
5
35
73
42
85
42
56
35
55
34
10
95:5
67:33
>95:5
67:33
>95:5
66:34
>95:5
75:25
60:40
1
2
3
4
10
20
70
68
80
>95:5
>95:5
>95.5
acetonitrile
acetonitrile
dimethoxyethane
dimethoxyethane
1,4-dioxane
1,4-dioxane
DMF
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.0 mmol), 1b (0.04-
0.2 mmol), sodium carbonate (1.5 mmol), diphenylsilane (1.1 mmol), and
toluene (3.0m). [b] Yields were determined by GCMS/MS against a cali-
brated internal standard (undecane) and were performed in duplicate.
[c] E/Z ratio was determined by GCMS/MS.
9
10
tert-butanol
5
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.0 mmol), 1b
(0.2 mmol), sodium carbonate (3.0 mmol), diphenylsilane (1.5 mmol), and
requisite solvent (3.0m). [b] Yields were determined by GCMS/MS
against a calibrated internal standard (undecane) and were performed in
duplicate. [c] E/Z ratio was determined by GCMS/MS.
insufficient rate of phosphane formation to effect the iso-
merization of the product after consumption of the aldehyde
(see further).
Following optimization of the base, an assessment of the
precatalyst loading was performed (Table 5). Pleasingly, the
results demonstrated that a loading of just 4 mol% was nec-
essary to achieve an acceptable yield (Table 5, entry 1). Im-
portantly, a loading of 4 mol% is often used in palladium
cross-coupling reactions and a comparable loading of an or-
ganocatalyst is impressive.[31] Indeed, there was no differ-
ence between loadings of 4 and 10 mol% (Table 5, entries 1
and 2, respectively).
Subsequent to protocol optimization, a substrate evalua-
tion was undertaken (Table 6) by utilizing aliphatic, aryl,
and heterocyclic aldehydes in combination with organoha-
lides containing ylide-stabilizing ester, nitrile, ketyl, and
electron-deficient benzyl moieties. Of particular note is the
use of heterocyclic aldehydes in the synthesis of 6, 7, 9, and
22–24. The synthesis of 23 is of specific significance because
this reaction was the first Z-selective catalytic Wittig reac-
Table 4. Temperature study.[a]
Entry
T [8C]
Yield [%][b]
E/Z[c]
66:34
66:34
66:34
1
2
3
4
60
70
80
33
49
62
60
100
>95:5
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.1 mmol), 1b
(0.1 mmol), sodium carbonate (1.5 mmol), diphenylsilane (1.5 mmol), and
toluene (3.0m). [b] Yields were determined by GCMS/MS against a cali-
brated internal standard (undecane) and were performed in duplicate.
[c] E/Z ratio was determined by GCMS/MS.
ꢀ
that the kinetic selectivity of the reaction is 66:34 and a
post-olefination isomerization occurs, thus converting the
Z olefin into an E olefin over time. We postulate that this
isomerization is phosphane mediated (Scheme 3). The lack
of a diastereoselective process when either triACHTNUTRGNEmNUG ethoxysilane
or temperatures lower than 1008C are employed (compare
tion. This Z selectivity may be because the N Ts group sta-
bilizes the formation of cis-oxaphosphetane, as observed by
Byrne and Gilheany.[12] Alternatively, steric influences could
explain the selectivity; nevertheless, this result offers key in-
sight toward the development of a Z-selective methodology.
In addition, the use of aliphatic aldehydes, in particular cit-
ronellal, gave olefins 12, 16, and 21 in good yields. The use
of electron-deficient benzyl bromides produce stilbene de-
rivatives 19 and 20 in good yields, but with moderate selec-
tivity (E/Z=60:40). To demonstrate the application of the
catalytic Wittig methodology further, 7 was produced on a
scale of 30 mmol (67% yield, E/Z>95:5, 48 h, 908C) using
just 4 mol% of 1b. Although diphenylsilane was the reduc-
ing agent of choice for the majority of substrates, trimethox-
ysilane was employed in situations in which the removal of
a siloxane byproduct was an issue, for example in the syn-
thesis of 22–24. These results neatly highlighted the benefit
in decoupling the terminal reducing agent from the olefinat-
ing agent because simply switching the silane compound
solves purification issues.
Table 2, entry 4 and Table 4, entries 1–3) may result from an
Scheme 3. Proposed phosphane-mediated isomerization of (Z)-4.
15284
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 15281 – 15289

Organocatalysis of the Wittig Reaction
FULL PAPER
Table 6. Substrate study using sodium carbonate for the CWR.[a]
than a,b-unsaturated products because propenenitrile ole-
fins do not undergo phosphane-mediated isomerization.
These findings may also have implications for the assess-
ment of diastereoselectivity in stoichiometric Wittig reac-
tions.
Although the results illustrate the considerable scope of
the catalytic Wittig reaction (Table 6), some difficulties were
noted during the application of the reaction. It was necessa-
ry to use finely ground sodium carbonate to achieve high
yields and to ensure rapid stirring throughout the reaction.
The use of a soluble base would alleviate this concern;
therefore, a series of soluble bases with a range of pKa
values (in water) for the conjugate acid between pKa 7 and
14 were screened for the CWR (Table 7). Of the soluble
Table 7. Soluble base study.[a]
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.1 mmol), 1b
(0.1 mmol), base (1.5 mmol), diphenylsilane (1.2 mmol), and toluene
(3.0m); the yields were determined by GCMS/MS against a calibrated in-
ternal standard (undecane) and were performed in duplicate.
[a] The compound number, yield, and E/Z ratio are given for each prod-
uct; the reactions were performed in duplicate (see the Supporting Infor-
mation for detail). [b] E/Z ratio was determined by 1H NMR spectro-
scopic analysis of the unpurified reaction mixture. [c] This reaction per-
formed with 4 mol% of 1b resulted in 73% yield. [d] This reaction per-
formed with methyl chloroacetate resulted in 68% yield. [e] A single re-
action with (MeO)3SiH (2.0 mmol) resulted in 85% yield. [f] This
reaction performed on a scale of 10.7 mmol with 10 mol% 1b resulted in
63% yield. [g] This reaction performed on a scale of 30 mmol with
4 mol% of 1b at 908C over 48 h resulted in 67% yield. [h] Performed
with (MeO)3SiH (2.0 mmol).
bases examined, only N,N-diisopropylethylamine (DIPEA;
Hꢂnigꢀs base) was shown to be suitable for use in the CWR.
The success of DIPEA hinged on two points: 1) the pKa
value of DIPEA-H+ is 11.4 (in water; 9.1 in DMSO), which
is ideal for the deprotonation of the phosphonium salt;
2) the steric bulk and lack of nucleophilicity of DIPEA en-
sured that unwanted side reactions did not occur.
Further examination of our substrate study revealed a
trend in diastereoselectivity, a,b-unsaturated products dem-
onstrated considerably higher E selectivity relative to prope-
nenitrile products (i.e., 5, 6, 11, and 12). To probe this differ-
ence in diastereoselectivity, we revisited the proposed phos-
phane-mediated isomerization event (Scheme 3) and carried
out an isomerization study with both (Z)-4 and (Z)-5
(Figure 2). The results were striking, after 10.5 hours full iso-
merization was observed for 4, whereas 5 showed only a
trace of isomerization (below the limit of quantitation).
Indeed, after 50 hours the conversion of (Z)-5 into (E)-5
had proceeded in just 6% conversion (see the Supporting
Information). In the absence of a phosphane, no isomeriza-
tion to (E)-4 occurred (see the Supporting Information).
These results point to the production of propenenitriles
(through bromoacetonitrile as the organohalide) as a better
probe for the selectivity of the catalytic Wittig system rather
After the identification of a soluble base, an assessment
of the phosphane oxide/phosphane structure was undertaken
(Table 8). Unsurprisingly, 1b and 1c were equally effective
in the CWR and could be used interchangeably, thus dem-
onstrating that the methyl group in 1b was a mere spectator
in the CWR. Additionally, the unreduced analogue of 1b,
that is, 1a, was also effective in the CWR, thus indicating
that reductive removal of the carbon–carbon double bond in
1a may have been unnecessary. Acyclic phosphanes, includ-
ing triphenylphosphane (TPP), were essentially ineffective
and resulted in <5% yield. The failure of acyclic phos-
phanes in the CWR stems from an insufficient rate of oxide
reduction and vindicates the initial decision to employ cyclic
phosphane oxides in the preliminary studies. Of the other
phosphane structures screened, only 5-phenyldibenzophos-
AHCTUNGTRENNUNG
Chem. Eur. J. 2013, 19, 15281 – 15289
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15285

C. J. OꢀBrien et al.
Figure 2. Isomerization of (Z)-4 and (Z)-5.
Table 8. Screening of phosphane catalysts.[a]
Table 9. Process-friendly solvents study.[a]
Entry
Solvent
Conversion [%][b]
E/Z[c,d]
1
2
3
4
5
6
7
toluene
cyclopentyl methyl ether
tert-butyl acetate
2-methyltetrahydrofuran
dimethyl carbonate
3-methyl-2-butanone
a,a,a-trifluorotoluene
96 (85)
96 (85)
94
93
92
72:28
69:31
69:31
69:31
70:30
70:30
69:31
82
80
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.2 mmol) 1b
(0.1 mmol), DIPEA (1.1 mmol), diphenylsilane (1.1 mmol), and the
requisite solvent (3.0m). [b] Conversions were determined by using
1H NMR spectroscopy, and selected examples were purified to determine
the yield of the isolated product (shown in parentheses). [c] E/Z ratio
was determined by 1H NMR spectroscopic analysis of the unpurified re-
action mixture. [d] When the reactions were performed with 1.2 equiva-
lents of diphenylsilane, which ensured phosphane predominated at the
end of the reaction, the E/Z selectivity was >95:5 as expected.
[a] Benzaldehyde (1.0 mmol), methyl bromoacetate (1.1 mmol), catalyst
(0.1 mmol), DIPEA (1.1 mmol), diphenylsilane (1.1 mmol), and toluene
(3.0m); the yields were determined by GCMS/MS against a calibrated in-
ternal standard (undecane) and were performed in duplicate. E/Z ratio
was determined by GCMS/MS. Cy=cyclohexyl. Cyp=cyclopentyl.
also influence the effectiveness of the phosphane-led
alkene-isomerization process, which would result in erosion
of the diastereoselectivity.
additional ring strain caused by fusing two of the aromatic
rings and, consequently, it proves to be more susceptible to
reduction than TPP oxide. The inferior yield and diastereo-
selectivity obtained by utilizing 27 relative to type 1 cata-
lysts probably results from the lower nucleophilicity of the
lone pair of electrons on the phosphorus atom to that of
1a–c. This effect stems from delocalization of the lone pair
of electrons in the corresponding phosphane of 27 within its
wider p system. Such a delocalization of the lone pair of
electrons on the phosphorus atom would impact the rate of
formation of the phosphonium salt and consequently effect
the yield of the olefin. Retardation in the nucleophilicity of
the lone pair of electrons on the phosphorus atom would
Following the modification of the CWR to accommodate
the use of a soluble base, the choice of the reaction solvent
was revisited, focusing on process-friendly solvents
(Table 9).[32] The solvent is well known to have an effect on
the diastereoselectivity of the Wittig reaction,[11] therefore
conditions were selected that prevented phosphane-mediat-
ed isomerization. Such conditions would allow our solvent
study to identify a deviation in yield and diastereoselectivity.
The results revealed that a high yield of the olefin was main-
tained across all the solvents tested (Table 9). Furthermore,
identical diastereoselectivity was also observed, thus indicat-
ing that the geometry of the olefin is governed by the phos-
15286
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 15281 – 15289

Organocatalysis of the Wittig Reaction
FULL PAPER
phane structure and not a solvent effect. These results dem-
onstrate that optimization of the phosphane/phosphane
oxide structure should produce a more selective process.
Employment of the DIPEA/CWR conditions broadened
the reaction scope to include furyl, indolyl, and pyridyl alde-
hydes (Table 10, 34, 39–41, 47, 49, and 50). In addition, the
DIPEA conditions were used to prepare a further 21 exam-
ples of trisubstituted olefins. A comparison of 1b and 1c
shows that the methyl group in 1b, as expected, does not
result in a significant difference in yield or selectivity (see
37–40, 42, and 47) and that these phosphane oxides may be
used interchangeably. Compounds 44–46 are of particular in-
terest because these are structurally similar to resveratrol,
and various methoxylated derivatives of resveratrol have
been shown to have anticancer activity.[33] The preparation
ꢀ
of 52, which features a free N H functional group, demon-
strates the mild nature of these reaction conditions. To high-
light the utility of the CWR further, its application in the
synthesis of the real-world example donepezil hydrochloride
was undertaken. Donepezil, under the trade name Aricept,
is used in the palliative treatment of mild-Alzheimer dis-
ease.[34] Donepezil hydrochloride is readily available from
precursor 54 through a hydrogenation reaction and HCl-salt
formation (Scheme 4). Compound 54 in turn is accessible
through a CWR. Pleasingly, 54 was easily prepared on a
multigram scale (12.2 g, 74%, E/Z>95:5) by utilizing a
10 mol% loading of 1c. When prepared on a small scale
Scheme 4. Retrosynthesis of donepezil hydrochloride and preparation of
54 by using CWR.
(1.0 mmol), this compound was obtained in 76% yield, thus
demonstrating that the scale has little variation on the yield.
Conclusion
We have developed the first catalytic (in phosphane) Wittig
reaction by utilizing an organosilane that chemoselectively
Table 10. Substrate study using DIPEA in the CWR.[a]
[a] The compound number, yield, E/Z ratio, phosphane oxide, and loading are given for each product. The reactions were performed in duplicate (see
1
the Supporting Information for details). [b] E/Z ratio was determined by H NMR spectroscopic analysis of the unpurified reaction mixture. [c] The reac-
tion was performed with (MeO)3SiH (2.0 mmol). [d] Reaction time was 12 h. [e] The aldehyde and organohalide were added together in 12 portions over
6 h. [f] The reaction was performed with phenylsilane in cyclopentyl methyl ether at 608C; only the E diastereomer was isolated. [g] The reaction was
performed with phenylsilane in ethyl acetate at 608C.
Chem. Eur. J. 2013, 19, 15281 – 15289
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15287

C. J. OꢀBrien et al.
10:90!0:100) to afford 54 as a brown solid (12.2 g, 74%, E/Z>95:5).
1H NMR (400 MHz, CDCl3, 208C): d=7.13–7.30 (m, 6H), 6.82 (brs,
1H), 6.58 (brd, J=9.6 Hz, 1H), 3.90 (s, 3H), 3.85 (s, 3H), 3.51 (d, J=
0.8 Hz, 2H), 3.46 (s, 2H), 2.86 (brd, J=11.6 Hz, 2H), 2.21–2.31 (m, 1H),
1.99 (td, J=11.2, 2.0 Hz, 2H), 1.50–1.65 ppm (m, 4H); 13C NMR
(100 MHz, CDCl3): d=192.7, 155.3, 149.5, 144.6, 139.9, 138.2, 135.7,
131.9, 129.3, 128.3, 127.1, 107.3, 105.1, 63.6, 56.3, 56.2, 53.1, 37.3, 31.3,
29.6 ppm; the spectral data was consistent with previous data.[34b]
reduces a phosphane oxide precatalyst to a phosphane. Fur-
thermore, to our knowledge, this is the only transition-/
heavy-metal-free catalytic olefination proACTHUNRTGNEUNGcess to date, ex-
cluding proton-catalyzed elimination reactions. The generali-
ty of this methodology is demonstrated in Tables 6 and 10,
with 23 examples prepared by using sodium carbonate and a
further 27 examples by using DIPEA. Both protocols can
employ aliphatic, aryl, and heterocyclic aldehydes to pro-
duce both di- and trisubstituted olefins in moderate-to-high
yields (60–96%). The CWR may be carried out on a large
scale, as highlighted in the synthesis of 54 on a 44 mmol
scale to yield 12.2 g (74%) of a precursor to the Alzheimer
drug donepezil hydrochloride. Kinetic E/Z selectivity was
generally 66:34; however, E selectivity could be achieved
through a phosphane-mediated isomerization event for a,b-
unsaturated products. Of note are the synthesis of 23 and
40, which are the first examples of Z-selective CWRs. These
results may offer valuable insight for further studies toward
the development of a Z-selective CWR. Further investiga-
tions toward lowering the reaction temperature and extend-
ing the methodology to other ylide types and carbonyl sub-
strates are on-going and will be reported in due course.
Acknowledgements
We thank Peakdale Molecular Ltd for the gift of heterocyclic aldehydes,
and Dr. M. Feeney (Trinity College Dublin) for high-resolution mass
spectrometry. Financial support for this work was received from The Uni-
versity of Texas at Arlington, Dublin City University (DCU, Career
Start), and Enterprise Ireland (EI, Grant No. CF/2011/1029).
[1] a) K. K.-C. Liu, J. Li, S. Sakya, Mini-Rev. Med. Chem. 2004, 4,
1105–1125; b) A. Saklani, S. K. Kutty, Drug Discovery Today 2008,
13, 161–171; c) K. C. Nicolaou, E. J. Sorensen, Classics in Total Syn-
thesis: Targets, Strategies, Methods, VCH, Weinheim, 1996; d) K. C.
Nicolaou, S. A. Snyder, Classics in Total Synthesis II: More Targets,
Strategies, Methods, Wiley-VCH, Weinheim, 2003.
[2] J. Clayden, N. Greeves, S. Warren, Organic Chemistry, 2nd ed.,
Oxford University Press, New York, 2012.
[3] L. Kꢂrti, B. Czakꢃ, Strategic Applications of Named Reactions in Or-
ganic Synthesis, Elsevier Academic Press, San Diego, 2005.
[4] a) G. Wittig, G. Geissler, Justus Liebigs Ann. Chem. 1953, 580, 44–
57; b) G. Wittig, U. Schollkopf, Chem. Ber. 1954, 87, 1318–1330.
[5] D. J. Peterson, J. Org. Chem. 1968, 33, 780–784.
[6] a) M. Julia, J.-M. Paris, Tetrahedron Lett. 1973, 14, 4833–4836;
b) P. J. Kocienski, B. Lythgoe, S. Ruston, J. Chem. Soc. Perkin Trans.
1 1978, 829–834.
[7] a) P. R. Blakemore, W. J. Cole, P. J. Kocienski, A. Morley, Synlett
1998, 26–28; b) J. B. Baudin, G. Hareau, S. A. Julia, Tetrahedron
Lett. 1991, 32, 1175–1178.
[8] a) N. Calderon, H. Y. Chen, K. W. Scott, Tetrahedron Lett. 1967, 8,
3327–3329; b) P. Schwab, R. H. Grubbs, J. W. Ziller, J. Am. Chem.
Soc. 1996, 118, 100–110; c) M. Scholl, S. Ding, C. W. Lee, R. H.
Grubbs, Org. Lett. 1999, 1, 953–956; d) S. B. Garber, S. R. Kings-
bury, B. L. Gray, A. H. Hoveyda, J. Am. Chem. Soc. 2000, 122,
8168–8179; e) J. A. Love, J. P. Morgan, T. M. Trnka, R. H. Grubbs,
Angew. Chem. 2002, 114, 4207–4209; Angew. Chem. Int. Ed. 2002,
41, 4035–4037; f) J. S. Murdzek, R. R. Schrock, Organometallics
1987, 6, 1373–1374; g) R. R. Schrock, Tetrahedron 1999, 55, 8141–
8153; h) K. C. Nicolaou, P. G. Bulger, D. Sarlah, Angew. Chem.
2005, 117, 4564–4601; Angew. Chem. Int. Ed. 2005, 44, 4490–4527.
[9] R. F. Heck, J. P. Nolley, Jr., J. Org. Chem. 1972, 37, 2320–2322.
[10] a) D. Edmonds, A. Abell in Modern Carbonyl Olefination (Ed.: T.
Takeda), Wiley-VCH, Weinheim, 2004, pp. 1–17; b) A. Abell,
D. M. K. Edmonds in Organophosphorus Reagents: A Practical Ap-
proach in Chemistry (Ed.: P. J. Murphy), Oxford University Press,
Oxford, 2004, pp. 99–127; c) O. I. Kolodiazhnyi, Phosphorous
Ylides: Chemistry and Application in Organic Synthesis, Wiley-
VCH, New York, 1999; d) K. C. Nicolaou, M. W. Hꢄrter, J. L. Gunz-
ner, A. Nadin, Liebigs Ann. Recl. 1997, 1283–1301; e) N. J. Law-
rence in Preparation of Alkenes: A Practical Approach (Ed.: J. M. J.
Williams), Oxford University Press, Oxford, 1995; f) B. E. Maryan-
off, A. B. Reitz, Chem. Rev. 1989, 89, 863–927.
Experimental Section
See the Supporting Information for the general experimental and full de-
tails of the syntheses and characterizations. Representative procedures
are given below.
General procedure for the CWR with sodium carbonate: A 1-dram vial
equipped with a stirring bar was charged with phosphane oxide 1b
(19 mg, 0.10 mmol, 10 mol%) and sodium carbonate (1.5 mmol,
1.5 equiv) in air. Other solid reagents were added to the mixture at this
point, such as the aldehyde (1.0 mmol, 1.0 equiv) and organohalide (1.3–
1.5 mmol, 1.3–1.5 equiv). The vial was sealed with a septum and purged
with argon. Toluene (0.33 mL) and liquid reagents were introduced into
the mixture, such as the aldehyde (1.0 mmol, 1.0 equiv), organohalide
(1.1–1.5 mmol, 1.1–1.5 equiv), and silane (1.1–2.0 mmol, 1.1–2.0 equiv).
The septum was replaced with a PTFE-lined screw cap under an inert at-
mosphere, and the reaction was heated at 1008C for 24 h. The crude reac-
tion mixture was filtered through a plug of Celite, concentrated in vacuo,
and purified by means of flash-column chromatography. Note that it is
important that the reactions are stirred vigorously to achieve the maxi-
mum yield.
General procedure for the CWR with DIPEA: A 1-dram vial equipped
with a stirring bar was charged with phosphane oxide 1b or 1c (0.04–
0.1 mmol, 4–10 mol%). Other solid reagents were added at this point,
such as the aldehyde (1.0 mmol, 1.0 equiv) and organohalide (1.1 mmol,
1.1 equiv). The vial was sealed with a septum and purged with argon. The
solvent (0.33–0.50 mL) and any liquid reagents were introduced to the
mixture, such as the aldehyde (1.0 mmol, 1.0 equiv), organohalide (1.1–
1.5 mmol, 1.1–1.5 equiv), DIPEA (1.1–1.3 mmol, 1.1–1.3 equiv), and
silane (1.1–2.0 mmol, 1.1–2.0 equiv). The septum was replaced with a
PTFE-lined screw cap under an inert atmosphere, and the reaction was
heated at 1008C for 24 h. The crude reaction mixture was concentrated
in vacuo and purified by means of flash-column chromatography.
Synthesis of 54 on a multigram scale: Compound 54 was prepared from
1-benzylpiperidine-4-carbaldehyde (8.9 g, 43.6 mmol, 1.0 equiv), 2-
bromo-5,6-dimethoxy-1-indanone (13.0 g, 48.0 mmol, 1.1 equiv), DIPEA
(8.4 mL, 48.0 mmol, 1.1 equiv), and diphenylsilane (9.7 mL, 52.3 mmol,
1.2 equiv) with 1c (785 mg, 4.3 mmol, 10 mol%) in toluene (14.4 mL) in
a pressure vessel (350 mL) at 1008C for 24 h. The crude product was pu-
rified by means of dry flash chromatography (hexane/ethyl acetate,
[11] For leading discussions on mechanisms and selectivity, see: a) R.
Robiette, J. Richardson, V. K. Aggarwal, J. N. Harvey, J. Am. Chem.
Soc. 2006, 128, 2394–2409; b) R. Robiette, J. Richardson, V. K. Ag-
garwal, J. N. Harvey, J. Am. Chem. Soc. 2005, 127, 13468–13469;
c) V. K. Aggarwal, J. R. Fulton, C. G. Sheldon, J. de Vicente, J. Am.
Chem. Soc. 2003, 125, 6034–6035; d) E. Vedejs, M. J. Peterso, in Ad-
vances in Carbanion Chemistry, Vol. 2 (Ed.: V. Snieckus), 1996,
15288
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 15281 – 15289

Organocatalysis of the Wittig Reaction
FULL PAPER
pp. 1–85; e) E. Vedejs, M. J. Peterson, Top. Stereochem. 1994, 21, 1–
157; f) for a review of asymmetric Wittig-type olefinations, see: T.
Rein, T. M. Pedersen, Synthesis 2002, 579–584; g) for stoichiometric
recycling of triphenylphosphane in an industrial setting, see: H.
Pommer, Angew. Chem. 1977, 89, 437–443; Angew. Chem. Int. Ed.
Engl. 1977, 16, 423–429.
[24] K. Naumann, G. Zon, K. Mislow, J. Am. Chem. Soc. 1969, 91, 7012–
7023.
[25] T. Imamoto, S. Kikuchi, T. Miura, Y. Wada, Org. Lett. 2000, 2, 87–
90.
[26] G. Keglevich, A.-C. Gaumont, J.-M. Denis, Heteroat. Chem. 2001,
12, 161–167.
[12] For a recent discussion on E/Z selectivity by employing b-heteroa-
tom-substituted aldehydes, see: A. Byrne, D. G. Gilheany, J. Am.
Chem. Soc. 2012, 134, 9225–9239.
[27] F. G. Mann, The Heterocyclic Derivatives of Phosphorus, Arsenic,
Antimony and Bismuth, Vol. 1, 2nd ed., Wiley-Interscience, London,
1970.
[13] a) P. Cao, C.-Y. Li, Y.-B. Kang, Z. Xie, X.-L. Sun, Y. Tang, J. Org.
Chem. 2007, 72, 6628–6630; b) S. Zhu, Y. Liao, S. Zhu, Org. Lett.
2004, 6, 377–380; c) W.-D. Dai, A. Wu, H. Wu, Tetrahedron: Asym-
metry 2002, 13, 2187–2191; d) L. Shi, W. Wang, Y. Wang, Y. Huang,
J. Org. Chem. 1989, 54, 2027–2028.
[14] Z.-Z. Huang, Y. Tang, J. Org. Chem. 2002, 67, 5320–5326.
[15] Y. Liao, Y.-Z. Huang, Tetrahedron Lett. 1990, 31, 5897–5900.
[16] J. A. Kerr in CRC Handbook of Chemistry and Physics, 81 st ed.
(Ed.: D. R. Lide), Boca Raton, 2000.
[17] F. E. Kuhn, A. M. Santos, Mini-Rev. Org. Chem. 2004, 1, 55–64.
[18] a) K. K. Kroening, M. J. Solivio, M. Garcꢅa-Lꢃpez, A. Puga, J. A.
Caruso, Metallomics 2009, 1, 59–66; b) Y. Hu, X. Jin, E. T. Snow,
Toxicol. Lett. 2002, 133, 33–45; c) J. Henriksson, A. Johannisson, P.-
A. Bergqvist, L. Norrgren, Arch. Environ. Contam. Toxicol. 1996,
30, 213–219.
[19] A. Taylor, Biol. Trace Elem. Res. 1996, 55, 231–239.
[20] K. A. Winship, Adverse Drug React. Acute Poisoning Rev. 1987, 6,
67–90.
[28] S. Dꢅez-Gonzꢆlez, S. P. Nolan, Org. Prep. Proced. Int. 2007, 39, 523–
559.
[29] G. Keglevich, M. Fekete, T. Chuluunbaatar, A. Dobꢃ, V. Harmat, L.
˝
Toke, J. Chem. Soc. Perkin Trans. 1 2000, 4451–4455.
[30] Although all the tested grades of sodium carbonate performed well,
high levels of transition-metal contamination of the carbonate base
should be avoided; for catalysis that employs transition-metal con-
taminants in carbonate bases, see: a) N. E. Leadbeater, M. Marco,
Angew. Chem. 2003, 115, 1445–1447; Angew. Chem. Int. Ed. 2003,
42, 1407–1409; b) N. E. Leadbeater, M. Marco, J. Org. Chem. 2003,
68, 5660–5667; c) N. E. Leadbeater, Nat. Chem. 2010, 2, 1007–1009.
[31] C. C. C. Johansson Seechurn, M. O. Kitchning, T. J. Colacot, V.
Snieckus, Angew. Chem. 2012, 124, 5150–5174; Angew. Chem. Int.
Ed. 2012, 51, 5062–5085.
[32] a) K. Alfonsi, J. Colberg, P. J. Dunn, T. Fevig, S. Jennings, T. A.
Johnson, H. P. Kleine, C. Knight, M. A. Nagy, D. A. Perry, M. Stefa-
niak, Green Chem. 2008, 10, 31–36; b) R. K. Henderson, C. Jimꢇ-
nez-Gonzꢆlez, D. J. C. Constable, S. R. Alston, G. C. A. Inglis, G.
Fisher, J. Sherwood, S. P. Binks, A. D. Curzons, Green Chem. 2011,
13, 854–862.
[21] a) P. A. Dasgupta, Talanta 2002, 58, 1–2; b) H.-Y. Chiou, S.-T.
Chiou, Y.-H. Hsu, Y.-L. Chou, C.-H. Tseng, M.-L. Wei, C.-J. Chen,
Am. J. Epidemiol. 2001, 153, 411–418; c) A. H. Welch, D. B. West-
john, D. R. Helsel, R. B. Wanty, Ground Water 2000, 38, 589–604.
[22] A. de Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling Reac-
tions, 2nd ed., Wiley-VCH, Weinheim, 2004.
[33] S. Fulda, Drug Discovery Today 2010, 15, 757–765.
[34] a) H. Sugimoto, Y. Iimura, Y. Yamanishi, K. Yamatsu, J. Med.
Chem. 1995, 38, 4821–4829; b) R. Janaki Rama Rao, A. K. S. Bhu-
jangaRao, Y. L. N. Murthy, Synth. Commun. 2007, 37, 2847–2853.
[23] a) C. J. OꢀBrien, J. L. Tellez, Z. S. Nixon, L. J. Kang, A. L. Carter,
S. R. Kunkel, K. C. Przeworski, G. A. Chass, Angew. Chem. 2009,
121, 6968–6971; Angew. Chem. Int. Ed. 2009, 48, 6836–6839; b) S. P.
Marsden, Nat. Chem. 2009, 1, 685–687; c) I. J. S. Fairlamb, ChemSu-
sChem 2009, 2, 1021–1024.
Received: April 16, 2013
Published online: September 25, 2013
Chem. Eur. J. 2013, 19, 15281 – 15289
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15289