Vilsmeier-Haack Reagents. Novel Electrophiles for the One-Step Formylation of O-Silylated Ethers to O-Formates

Article abstract of DOI:10.1055/s-2004-815435

Various O-silylated substrates were effectively converted in one-step to their corresponding O-formates using electrophilic racemic and homochiral Vilsmeier-Haack reagents. Reactivity trends of these transformations were examined that, specifically, empha

Full text of DOI:10.1055/s-2004-815435

564
NEW TOOLS IN SYNTHESIS
Vilsmeier–Haack Reagents. Novel Electrophiles for the One-Step Formylation
of O-Silylated Ethers to O-Formates
????? ean-Paul Lellouche,* Vadim Kotlyar
Department of Chemistry, Bar-Ilan University, Ramat-Gan, 52900 Israel
Fax +972(3)5351250; E-mail: lellouj@mail.biu.ac.il
Received 30 October 2003
mediated by electrophilic Vilsmeier–Haack (VH) re-
agents. O-Formates can be removed in very mild condi-
tions (0.6 M NH4OH in MeOH, 20 °C) without affecting
sterically more hindered O-acetates, as well as hydrolyti-
cally sensitive O-silyl ethers.1,2 Typically, VH-reagents 1
(equilibrium mixture of the two salts 2 and 3) and 4 were
found effective for such Pg interconversions (Scheme 1).
VH reagents are electrophilic species that are able to
formylate nucleophilic substrates such as activated aro-
matics, double bonds, enolizable ketones, or amines.8,9
They have also found extensive use in halogenating and
dehydrating processes. Practically, these electrophiles can
be easily prepared by reaction of an inorganic acid halide
or anhydride [e.g. POCl3, SOCl2, COCl2, or (CF3SO2)2O]
with a N,N-disubstituted formylamide such as DMF at 0–
20 °C.
Abstract: Various O-silylated substrates were effectively convert-
ed in one-step to their corresponding O-formates using electrophilic
racemic and homochiral Vilsmeier–Haack reagents. Reactivity
trends of these transformations were examined that, specifically,
emphasized their synthetic potential.
Key words: Vilsmeier–Haack complexes, deprotection of O-silyl
ethers, one-step deprotection, preparation of O-formates, formyl
glycosides
Introduction
Synthetic planning toward complex multifunctional or-
ganic compounds emphasize the importance of a judi-
cious choice of specific protective groups (Pgs).1,2 Their
careful selection, their introduction, and cleavage condi-
tions are essential milestones critical for success. Not
astonishingly, current trends in the field are aimed at the
discovery of new Pgs or milder cleavage conditions. In
this regard, organic chemists devote much attention to
synthetic transformations that allowed Pg exchanges
without isolating deprotected intermediates. Such one-
step Pg interconversions should result in higher overall
yields, should bypass problematic intermediate deprotec-
tions, and should minimize waste during multi-step syn-
thesis. In this context, O-silyl ethers R-OSiR¢3 have
proved very popular as OH-protecting groups since (i)
they are easily introduced in high-yields, (ii) present mod-
ulatory chemical stabilities over a wide range of condi-
tions, and (iii) can be deprotected by fluoride-based,
acidic/basic or strongly reducing/oxidizing reagents. Nev-
ertheless, their deprotections can be very often problemat-
ic for sensitive multi-functional substrates,3,4 so alternate
orthogonal milder cleavage conditions involving Pg ex-
changes could be of major interest. Prior to our work, few
one-step transformations of O-silyl ethers to acetates5 or
benzyl/methyl,6 and THP ethers7 were known. Synthetic
limitations were clearly encountered due to harsh reaction
conditions (strong Lewis acids, in situ generated basic
alkoxides, high temperatures).
Scheme 1 VH-Reagents 1 and 4
O-TES and O-TBDMS Silylated Ethers
Early exploratory works dealt with reactivity studies of
various O-TES and O-TBDMS silyl ethers that were re-
acted with the prototypical VH-reagent 1 prepared from
POCl3 (Table 1, see also the Typical Experimental Proce-
dure).10 Substrates presenting different degrees of substi-
tution (primary versus secondary ethers), chemical type
(aliphatic, neopentyl, allylic, and propargylic ethers), and
multifunctionality (1,2-/1,4-diol ethers) were examined.
The formation of the corresponding O-formates by Pg ex-
change appeared very clean based on TLC and 1H-NMR
checking. Conversion yields were consistently in the me-
dium to the high range (60–98 %) although limited by the
volatility of formates 7–10 (Table 1, entries 3–6). Propar-
gylic and E-ethylenic functions were well tolerated in
these conditions (entries 3,4) while cis/trans-diastereo-
selectivities in 1,2-cyclohexane-di-O-formates 12/13 and
14/15 were retained when compared to the starting O-sil-
ylated diols (entries 8,9). Retention of configuration for
This ‘New Tool’ article will survey our most recent find-
ings dealing with the one-step conversion of O-silyl ethers
to their corresponding O-formates R-OSiR¢3 → R-OCHO
SYNLETT 2004, No. 3, pp 0564–0571
1
8.
0
2.
2
0
0
4
Advanced online publication: 26.01.2004
DOI: 10.1055/s-2004-815435; Art ID: T02003ST
© Georg Thieme Verlag Stuttgart · New York

NEW TOOLS IN SYNTHESIS
Novel Electrophiles for the One-Step Formylation of O-Silylated Ethers
565
substrates 10 and 11 has been observed, which is compat-
ible with the proposed formylation mechanism
(Scheme 3).10
O-Silylated D-Glucal Derivatives
The scope and limitations of the above Pg interconversion
mediated by the VH-reagent 1 needed further investiga-
tion, mainly considering multifunctional O-silyl ethers.11
O-Silylated
D-glucal-based
precursors
16–1911
(Scheme 2) with various degrees of steric hindrance at the
silicon were tested not only with reagent 1, but also with
the known more electrophilic VH-reagent 4 (prepared
from triflic anhydride Tf2O and DMF).12 These sugar-de-
rived cyclic enol ethers are acid-sensitive, contain a dense
array of O-silyl ethers of primary, secondary allylic nature
in a 1,2/1,3/1,4-relationships that should provide different
reactivities. The intracyclic enol ether itself could be com-
petitively C-formylated by reagents 1 and 4 at position C2
to afford C2-formylated glucal derivatives. For example,
the tri-O-benzyl-D-glucal is known to react in this way
with reagent 1.13
Scheme 2 (a) POCl3·DMF 1 or (CF3SO2)2O·DMF 4, anhyd DMF,
0–20 °C
Relative to these D-glucal-based precursors, reagent 1 and
the Tf2O-based reagent 4 were found able to formylate
substrates 17, 18, and 20–23 affording the corresponding
primary C6-O-formates 17a, 18a, and 20a in medium to
high yield (50–91%, Table 2). Globally, reagent 4 was
found to be more reactive and more efficient than reagent
1 resulting in shorter reaction times. As rationalized later
(Scheme 3), no competitive C2-formylation products
Biographical Sketches
Jean-Paul Lellouche was moved to Negev, the Insti- tions. His additional main
born in Constantine (Alge- tute for Applied Biosciences scientific contributions deal
ria) in 1955. He received his at the Ben-Gurion Universi- with asymmetric synthesis
diploma of Engineer in Or- ty (Beer-Sheva, Israel) as an using homochiral h4-dienyl
ganic Chemistry from the Associate-Professor in Or- tricarbonyliron(0) complex-
ESCIL (Ecole Supérieure de ganic Chemistry, where he es and related cations to-
Chimie Industrielle de Ly- received a Guastella fellow- ward biologically active di-
on, Claude Bernard Univer- ship from the Rashi Founda- HETEs,
sulfido-leuko-
sity, Lyon, France) in 1978. tion. He subsequently trienes, and phospholipids.
Thereafter, he joined the joined the Department of He has authored 68 papers,
CEA (Commissariat à l’En- Chemistry at the Bar-Ilan 5 patents and 2 book chap-
ergie Atomique, Gif-sur- University in October 2000. ters in organometallic and
Yvette, France) to earn his He is currently developing fluorine chemistries (see
PhD degree in 1981 under novel pyrrole-/carbazole- the following URL:
the direction of Dr. Louis based conducting polymers http://www.biu.ac.il/ESC/
Pichat (multi-step syntheses and related nanosized com- ch/faculty/lellouche/
of 14C-/3H-labeled organic posites for diverse bio- and jplellint.html).
molecules). In 1997, he immunosensing
applica-
Vadim Kotlyar was born in Pharmacology at the Chaim working on the develop-
Ekaterinenburg, Russia in Sheba Medical Center (Tel- ment of selective asymmet-
1974. He received his BSc Hashomer, Israel) as an ana- ric one-step conversions of
and MSc degrees in Phar- lytical chemist from 1997– O-silyl ethers to their corre-
maceutical Chemistry from 2000. After which, he sponding O-formates. His
the Bar-Ilan University (Ra- moved to the Department of research interests are in the
mat-Gan, Israel) in 1996 Chemistry of the Bar-Ilan areas of high-throughput ex-
and 1998, respectively. He University working with perimentation and asym-
performed his military ser- Professor Jean-Paul Lel- metric synthesis.
vice at the Institute of Clini- louche to complete his PhD
cal
Toxicology
and studies. He is currently
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York

566
J.-P. Lellouche, V. Kotlyar
NEW TOOLS IN SYNTHESIS
Table 1 Formylation of O-TES and O-TBDMS Silylated Ethers
Mediated by the VH-Reagent 1
TIPS >> TBDPS and 4: TES >> TBDMS >> TBDPS ca.
TIPS. As expected, this reactivity scale follows increasing
steric hindrance at the silicon.
Entry O-Silylated Ether
O-Formatec Yield Time
(%)
(h)
Nevertheless, some limitations have been already identi-
fied. For example, the tri-O-TES-D-glucal 16 was rapidly
destroyed (30 min reaction, results not shown in Table 2)
even in the presence of pyridine, while the dibenzylated
mono-O-TES-D-glucal 20 has been successfully formy-
lated by reagents 1 and 4 (entry 5, 70% and 84%). A ster-
ically hindered O-TIPS group could be formylated but the
formylation outcome depended on substrates. Compared
to the unreactive per-O-TIPS precursor 19, the less en-
cumbered and/or more electronically activated di-O-ben-
zyl-D-glucal 23 was successfully C(6)-O-formylated
(entries 4 and 8, 10 versus 85% and 91%).
1
5a
6a
91
4
5a
2
98
5
6a
3
7a
8a
62
69
5
4
7a
4
8a
Table 2 Selective O-Formylation of O-Silylated D-Glucal Deriva-
tives
5
9a
60
78
5
Entry Pre-
cursor
Product (R3)
1
4
9a
Yield Time
Yield Time
6
10a
14
(%)
70
80a
50
10
70
82
60
85
(h)
(%)
69
78a
65
10
84
85
70
91
(h)
1
2
3
4
5
6
7
8
17
17a (TBDMS)
8
6
17
18
19
20
21
22
23
17a (TBDMS)
72
8
48
8
10a
7
11a
88
3
18a (TBDPS)
19a (TIPS)
48
3
48
2
20a (Bn)
“
“
“
4.5
3.5
11a
24
8
8
8
8
12a
12a
79
74
14
14
a Anhyd pyridine was added (3.3 equiv/POCl3).
12b
13a
9
14a
14a
98
71
14
14
Conversion Mechanism
A likely mechanism for these conversions has been exem-
plified for D-glucal-based precursors (Scheme 3). The
first is conversion of the sterically-driven electrophilic ad-
dition of VH-reagents 1 and 4 to the primary O-silyl ether
function affording intermediate oxonium cations of type
24. Favored by the formation of the thermodynamically
strong Si–Cl/Si–O bonds (111.0 and 128.2 kcal, respec-
tively),14 the b-elimination of the R¢3Si-X species affords
a positively charged C(6)-located imidate 25 that, elec-
tronically deactivates additional O–Si bonds to react sim-
ilarly. This explains why a three/five-fold excess of
reagents 1 or 4 did not result in the isolation of some ac-
companying di/tri-formylated products (results not shown
in Table 2).15,16 Smooth hydrolysis of imidates of type 25
produces the corresponding C(6)-O-formates.
14b
15a
a R1 = TBDMS.
b R1 = TES.
c R1 = CHO.
were ever isolated during these trials. Whatever reagent,
anhydrous pyridine was found to be compatible with the
transformation. In this case, conversion yields were
slightly increased at the expense of reaction times (entries
1 and 2). This last result is particularly important regard-
ing acid-sensitive substrates. The silyl groups in 20–23
could be graded by their decreasing O(6)-formylation re-
activity toward 1 and 4 as follows: 1: TES >> TBDMS >>
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York

NEW TOOLS IN SYNTHESIS
Novel Electrophiles for the One-Step Formylation of O-Silylated Ethers
567
the silyl group hindrance and on the number of VH-re-
agent equivalents (Table 3, entries 1 and 2, 2 and 14 h
reaction time, 94% and 85%). Whatever conditions, the 1-
O-TBDPS and 1-O-TIPS derivatives 28 and 29 were
found to be too hindered to react efficiently (entries 3 and
4, conversion yields less than 28 %, 14 h reaction time).
C2-Symmetrical Dialkoxysilanes R1O-Si(R2)2-OR1
Scheme 3 Imidate-related conversion mechanism
Intramolecular reactions involving species pre-organized
by linkage to a removable ‘temporary silicon connection’
have been recently developed to predictably enhance at-
tachment regio- and stereoselectively.18 Intramolecular
glycosidations, [4+2]-Diels–Alder condensations, ring-
closure metatheses, ene-ene/ene-yne [2+2]- and [3+2]-ni-
trone cycloadditions typically resulted in cyclic O-silylat-
ed dialkoxysilanes. Following condensation, the silicon
tether needed to be eliminated to liberate OH-containing
reaction products. This overall concept of ‘reacting pre-
organized species’ would be synthetically much more at-
tractive if a non-fluoride-based deprotective protocol
could be made available to liberate O-protected rather
than OH-products. Consequently, we examined the reac-
tivity of reagents 1 and 4 with model C2-symmetrical di-
alkoxysilanes R1O-Si(R2)2-OR1 31–36, that were easily
accessible from (–)-menthol and 3b-cholesterol respec-
tively (Scheme 5 and Table 4).19
1-O-Silylated D-Mannofuranose Derivatives
Reagents 1 and 4 were further tested on 1-O-silylated D-
mannofuranose derivatives 26–29. Corresponding 1-O-
formates should be produced via intermediate anomeric
imidates (Scheme 4 and Table 3).17 The rationale that sup-
ported these trials was to test the mild generation of posi-
tively charged anomeric imidates without any anomeric
activation by heavy toxic metals or strong Lewis acids.
Such transformations should be potentially useful for the
development of a general glycosidation protocol of ther-
modynamically stable 1-O-silylated donors, that should
be mediated by Vilsmeier–Haack reagents.
Scheme 4 Reaction of 1-O-silylated D-mannofuranose derivatives
26–29 with VH-reagents 1 and 4
Basically, reagents 1 and 4 were found to react quite dif-
ferently with 1-O-silylated precursors 26–29 that pos-
sessed increasing steric requirements at silicon. Reagent 1
reacted with 26 (a/b = 55:45, 3 h) and 27 (a/b = 95:5, 14
h) to afford the chromatographically labile a-chloride 30
(65–75 %). Most likely, anomeric oxonium or imidates
species could be substituted by nucleophilic chloride an-
ions generated in the medium. On the contrary, reagent 4,
that possesses a non-nucleophilic triflate counter-anion,
lead to the expected a-formate 26a in yields depending on
Scheme 5 Reaction of VH-reagents 1 and 4 with C2-symmetrical
di-alkoxysilanes
Whatever the conditions, reagent 1 was always found to
be the most efficient affording formates 10a and 11a in a
medium to good yield (46–83 %). Except for the less hin-
dered methyl-substituted dialkoxysilane 34 (R2 = Me, en-
try 10), the more electrophilic triflate-based reagent 4
invariably afforded low to average conversion yields (20–
68 %). This last result did not follow the usual efficiency
trend that characterized the O-formylation of simple O-si-
lylated ethers. Some other valuable features should be no-
ticed regarding reagent 1. Except for one entry (entry 9),
conversion yields were similar for the two reagent/sub-
Table 3 O-Formylation of Anomeric O-Silylated D-Mannofuranose
Derivatives Mediated by the VH-Reagent 4 (yields reported for 26a)
Entry Precursor
(R1)
4
4
4
(1.0 equiv) (2.0 equiv) (4.0 equiv)
1
2
3
4
26 (TES)
55
26
5
94
54
10
10
–
27 (TBDMS)
28 (TBDPS)
29 (TIPS)
85
28
15
< 5
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York

568
J.-P. Lellouche, V. Kotlyar
NEW TOOLS IN SYNTHESIS
Table 4 One-Step O-Formylation of C2-Symmetrical Dialkoxysilanes Mediated by VH-Reagents 1 and 4
Entry
Precursor (1.0 equiv)
1 or 4 (equiv)
O-Formate 10a
1 (%)
O-Formate 11a
1 (%)
4 (%)
20
30
36
68
56
66
–
4 (%)
–
1
2
31
32
33
31
32
33
34
35
36
34
35
36
1.0
“
73
80
80
82
77
82
–
–
–
–
3
“
–
–
4
2.0
“
–
–
5
–
–
6
“
–
–
7
1.0
“
82
75
46
75
76
83
50
50
30
78
59
5
8
–
–
9
“
–
–
10
11
12
2.0
“
–
–
–
–
“
–
–
a Reagents 1 and 4 have been reacted with C2-symmetrical dialkoxysilanes 31–36 in anhydrous DMF (–5 °C to r.t., 18 h) using the indicated
molar ratios between VH-reagents and substrates
strate ratios of 1.0 and 2.0 (entries 1–3 versus 4–6 and en-
tries 7 and 8 versus 10–12). Most likely, another reagent
1-like compound has been generated that possessed simi-
lar O-formylating capabilities. Remarkably, conversions
were found to be independent of increasing steric hin-
drance at silicon (R2 varying from Me to i-Pr).
Based on our former mechanism (Scheme 3), the overall
difference of reactivity observed between reagents 1 and
4 can be rationalized in the following simple way
(Scheme 6). The electrophilic addition of reagents 1 or 4
(1 equiv) to the first O–Si bond of dialkoxysilanes sub-
strates 31–36 should afford the imidate 42 (1 equiv) in ad-
dition to the neutral species 37 (X = Cl or Tf). The second
O–Si bond of this last silylated intermediate 37 can react
similarly, affording a second equivalent of 42 accompa-
nied by the dichlorosilane 40 or by the bistriflated silane
41. The critical step 37 → 39 should be clearly disfavored
when considering a less nucleophilic intermediate 37 (X =
OTf) versus 37 (X = Cl). Supporting this stepwise mech-
anism, the chromatographically labile silanol ether 38 re-
sulting from the hydrolysis of 37 (entry 6) has been
isolated and partially characterized by spectroscopic and
Scheme 6 (a) VH-reagents 1 or 4, anhyd DMF, 0–20 °C
Improved Solvent-Dependant One-Step O-Formyla-
tions. CH2Cl2/DCE- versus DMF-Based Protocols
Experiments described up to now employed DMF not
only as the N-formyl dialkylamine precursor of VH-re-
agents 1 and 4, but also as the conversion solvent (DMF-
based protocol). In a further step, various solvents ranging
from aprotic non-polar to polar were screened to find suit-
able candidates that should be compatible with one-step
O-formylations mediated by VH-reagents 1 and 4 (data
not shown). Similar conversions could be more efficiently
performed in anhydrous dichloromethane (CH2Cl2-based
protocol) or 1,2-dichloroethane (DCE-based protocol),
where DMF acted uniquely as a reagent. For comparison,
best results of O-formylations using the DMF-based pro-
tocol on the various O-silylated substrates 17–19
(Scheme 2), 32, 33 (Scheme 5) and 43–45 (Scheme 7)
were gathered in Table 5 (5th column). In parallel, the
same series of substrates was reacted using modified
1
analytical means [high-field H NMR, 300 MHz, CDCl3
and EI-MS/DCI-MS (NH3 and CH4)].
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York

NEW TOOLS IN SYNTHESIS
Novel Electrophiles for the One-Step Formylation of O-Silylated Ethers
569
Novel Racemic and Homochiral VH-Reagents
Interestingly, the two N-formyl dialkylamine and inorgan-
ic acid halide/anhydride components can be systematical-
ly changed to introduce molecular diversity. On the one
hand, novel electrophilic VH-reagents, different from the
former reagents 1 and 4, could be synthesized and
screened regarding their O-formylation capabilities. On
the other hand, CH2Cl2/DCE-based O-formylations are
even more attractive synthetically, since homochiral N-
formyl amines could afford novel homochiral VH-re-
agents. These optically active VH-reagents should pave
the way to innovative kinetic resolutions or deracemiza-
tions of racemic O-silylated substrates.
Scheme 7 Reactivity of VH-reagents 1 and 4 in CH2Cl2 or DCE
CH2Cl2/DCE-based protocols (6th column, see also the
Improved CH2Cl2/DCE-Based Experimental Procedure).
Depending on substrates, CH2Cl2- or DCE-based O-
formylations were generally found to be higher yielding
(Table 5, 71–98%) than DMF-based ones (10–82%). For
example, the strongly hindered per-O-TIPS D-glucal de-
rivative 19 could not be formylated easily by reagents 1
and 4 in DMF (conversion yields less than 10%, entry 5).
On the contrary, the DCE-based protocol afforded the cor-
responding primary C(6)-O-formate 19a in quite a high
86% yield (entry 6). Similar reactivity trends were ob-
served with the two O-TBDMS and O-TIPS glycols 44
and 45 toward primary O-formates 44a and 45a (entries
11 versus 12: 63% versus 90% yield, entry 13: 10% versus
71% yield). Again since useful for acid-sensitive sub-
strates, pyridine was found to be compatible with transfor-
mations operated in CH2Cl2/DCE (entries 1–4). Better
conversion yields for substrates 17 and 18 have been ob-
served (entries 1,2: 80% versus 89% in a shorter reaction
time: 72 h versus 14 h; entries 3,4: 65% versus 78%).
Table 5 O-Formylations in CH2Cl2 or DCE. Comparison with the
Standard DMF-Based Protocol
Entry
Sub-
strate
Prod-
uct
Time
(h)
1 and/or 4
1 and/or 4
DMF (%)a
Solvent (%)b
1
2
17
“
17a
“
72
14
“
1 (80)c
–
Scheme 8 Novel racemic and homochiral VH reagents 46–63
–
4: DCE (89)c
3
18
“
18a
“
4 (65)
–
Fulfilling the above first step, recent results from our lab-
oratory demonstrated unambiguously that additional race-
mic and homochiral VH-reagents could perform similar
O-formylations in DMF or DCE (Scheme 8 and Table 6).
New racemic VH-reagents that were included in this
preliminary study were: (COCl)2·DMF (46),20 MsCl·DMF
(47),21 TsCl·DMF (48),21 SO2Cl2·DMF (49),8,9
ClCO2Me·DMF (50),8,9 ClCO2Et·DMF (51),8,9 TM-
SOTf·DMF (52),22 NBS-PPh3·DMF (53),23 [Cl2·PPh3]
·DMF (54),24 [PhCOCl–AgOTf]·DMF (55),25 and, finally,
the last two C3Cl3N3·DMF (56) and Cl6N3P3·DMF (57),
formed by reaction of cyanuric chloride (C3Cl3P3, 64) and
2,4,6-trichloro[1,3,5]triazine (Cl6N3P3, 65) respectively
with DMF.26 The novel homochiral VH-reagents 58–63
[X and Y = Cl, Cl2P(O)O, OTf, O-heterocycle] were sim-
ilarly prepared in DCE by reaction of POCl3, Tf2O,
(COCl)2, SOCl2, C3Cl3N3 (64) and Cl6N3P3 (65) with the
commercially available homochiral N-formyl amine (S)-
2-methoxymethyl-pyrrolidine-1-carbaldehyde (66) [(S)-
MMPC, Aldrich, ee ≥ 99 %].
4
72
48
14
“
–
4: CH2Cl2 (78)c
–
5
19
“
19a
“
1 and 4 (10)
6
–
1: DCE (86)
1: CH2Cl2 (98)
1: CH2Cl2 (98)
–
7
32
33
43
“
10a
“
1 (77)
1 (82)
1 (61)c
–
8
“
9
43a
“
“
10
11
12
13
“
1: DCE (81)
–
44
“
44a
“
“
1 (63)
–
4
1: DCE (90)
45
45a
14
1 and 4 (10) 4: DCE (71)
a Results obtained with the DMF-based protocol.
b Results obtained with the CH2Cl2/DCE-based protocols.
c Anhyd pyridine was added.
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York

570
J.-P. Lellouche, V. Kotlyar
NEW TOOLS IN SYNTHESIS
DMF, 0 °C). The medium is then stirred at 20 °C until the reaction
is complete (TLC monitoring, reaction times indicated in the corre-
sponding Tables). After medium hydrolysis at 0 °C (sat. aq solution
of NaHCO3, 30 mL) and usual work-up, resulting crude O-formates
are purified by preparative flash chromatography on silica gel Mer-
ck (45–60 mm) to afford the corresponding pure compounds (reac-
tion times and conversion yields are indicated in the related Tables).
The O-formylation capabilities of these novel VH-re-
agents were screened using the test O-formylation of the
neopentyl rac-O-TES menthol rac-10 → rac-10a
(Table 1, entry 6). VH-reagents and solvent conditions
that provided a quantitative conversion (yields in rac-10a
better than 95 %) were identified and gathered in Table 6.
Interestingly, 7 new racemic (entries 1–7) and 4 homo-
chiral VH-reagents (entries 8–11) have been found to be
effective, emphasizing the generality and the synthetic
potential of such one-step O-formylations.
An Improved CH2Cl2/DCE-Based Experimental Procedure
POCl3 or Tf2O (1.1 mmol each O–Si bond in substrates) is dissolved
in anhyd CH2Cl2 or ClCH2CH2Cl containing anhyd DMF under ni-
trogen (CH2Cl2 or DCE/DMF: 2.0 mL/4.0 mmol, 320.0 mL), cooled
at –15 °C, stirred for 30 min at the same temperature and added
dropwise to a cold solution of O-silylated substrates (1.0 mmol, 1.0
mL of anhyd CH2Cl2 or DCE and 0.7 mL washing, 0 °C). The me-
dium is then stirred at 20 °C for the indicated times (TLC monitor-
ing). For trials including freshly distilled anhyd pyridine (2.5 mmol
per O–Si bond, 200.0 mL), the base is added to the O-silylated sub-
strate. After medium hydrolysis at 0 °C (biphasic mixture: sat. aq
solution of NaHCO3/Et2O 10 mL/10 mL) and usual aqueous work-
up, resulting crude compounds are purified by preparative flash
chromatography on silica gel Merck (45–60 mm) to afford the cor-
responding pure O-formates (reaction times and conversion yields
are indicated in Table 5).
Table 6 O-Formylation of rac-O-TES Menthol. Reactivity
Screening Using Racemic and Homochiral VH Reagents 46–63
Entry VH-Reagent
Solvent
1
2
(COCl)2·DMF
46
47
48
49
55
56
57
58
60
61
63
DMF and DCE
MsCl·DMF
DMF
3
TsCl·DMF
“
4
SO2Cl2·DMF
DCE
5
[PhCOCl-AgOTf] ·DMF
C3Cl3P3·DMF
DMF and DCE
Acknowledgement
6
“
The Bar-Ilan University and the Israeli Ministry of Immigrant
Absorption are acknowledged for their financial support (internal
research and laboratory equipment grants). The Rashi Foundation is
thanked for its financial support to JPL (funding of a Guastella
fellowship).
7
Cl6N3P3·DMF
“
8
POCl3·(S)-MMPC
(COCl)2·(S)-MMPC
SO2Cl2·(S)-MMPC
Cl6N3P3·(S)-MMPC
DCE
9
“
“
“
10
11
References
(1) Greene, T. W.; Wuts, P. G. M. Protective Groups in Organic
Synthesis, 3rd Ed.; John Wiley and Sons Inc.: New York,
1999.
(2) Kocieski, P. J. Protecting Groups; Georg Thieme Verlag:
Stuttgart, 1994.
Conclusions and Outlook
One-step O-formylations of diverse O-silylated substrates
have been mediated by electrophilic Vilsmeier–Haack re-
agents. These Pg interconversions have been found to be
general and synthetically efficient. Minimizing steps and
waste, they should provide a very stimulating entry to al-
ternative, more effective protective-deprotective strate-
gies useful in multi-step synthesis. On-going progress,
that should broaden their synthetic potential, will involve
the development of novel glycosidation reactions using
stable 1-O-silylated donors, and of new solid phase-sup-
ported VH-reagents for O-silylated ether scavenging
(scavenging polymer-supported VH-reagents). In paral-
lel, O-formylations mediated by homochiral VH-reagents
are currently developed targeting the kinetic resolution of
secondary/tertiary O-silylated ethers and the deracemiza-
tion of various meso O-silylated substrates. These works
will be reported in due time.
(3) Prakash, C.; Saleh, S.; Blair, I. A. Tetrahedron Lett. 1994,
35, 7565.
(4) Maiti, G.; Roy, S. C. Tetrahedron Lett. 1997, 38, 495.
(5) (a) Kim, S.; Lee, W. J. Synth. Commun. 1986, 16, 659.
(b) Oriyama, T.; Oda, M.; Gono, J.; Koga, G. Tetrahedron
Lett. 1994, 35, 2027. (c) Ganem, B.; Small, R. V. J. Org.
Chem. 1974, 39, 3728. (d) Oriyama, T.; Noda, K.; Yatabe,
K. Synlett 1997, 701. (e) Chandra, K. L.; Saravanan, P.;
Singh, V. K. Tetrahedron Lett. 2001, 42, 5309. (f) Sharma,
G. V. M.; Mahalingam, A. K.; Nagarajan, M.; Ilangovan, A.;
Radhakrishna, P. Synlett 1999, 1200. (g) Reddy, C. S.;
Smitha, G.; Chandrasekhar, S. Tetrahedron Lett. 2003, 44,
4693.
(6) Saunders, D. G. Synthesis 1988, 377.
(7) Suzuki, T.; Oriyama, T. Synthesis 2001, 555.
(8) Marson, C. M. Tetrahedron 1992, 48, 3659.
(9) Marson, C. M.; Giles, P. R. Synthesis Using Vilsmeier
Reagents; CRC Press: Boca Raton, 1994.
(10) Koeller, S.; Lellouche, J.-P. Tetrahedron Lett. 1999, 40,
7043.
(11) Lellouche, J.-P.; Koeller, S. J. Org. Chem. 2001, 66, 693.
(12) Martinez, A. G.; Alvarez, R. M.; Barcina, J. O.; de la Moya
Cerero, S.; Vilar, E. T.; Fraile, A. G.; Hanack, M.;
Subramanian, L. R. J. Chem. Soc., Chem. Commun. 1990,
1571.
Formylation; General Procedure
POCl3 (1.1 equiv per O–Si bond to be potentially formylated) is dis-
solved in anhyd DMF (3 mL) under nitrogen, cooled to 0 °C, agitat-
ed for 30 min at the same temperature and added dropwise to a cold
solution of the O-silylated substrate (1.0 mmol, 2 mL of anhyd
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York

NEW TOOLS IN SYNTHESIS
Novel Electrophiles for the One-Step Formylation of O-Silylated Ethers
571
(13) Ramesh, N. G.; Balasubramanian, K. K. Tetrahedron Lett.
1991, 32, 3875.
(14) Lide, D. R. Handbook of Chemistry and Physics, 8th Ed.;
CRC Press: Boca Raton, 1997, 9/667.
(15) Friesen, R. W.; Danishefsky, S. J. J. Am. Chem. Soc. 1989,
11, 6656.
(16) Danishefsky, S. J.; Bilodeau, M. T. Angew. Chem., Int. Ed.
Engl. 1996, 35, 1381.
(17) Cohen, Y. MSc Thesis; Bar-Ilan University: Ramat-Gan
Israel, 2002.
(18) (a) Bols, M.; Skrydstrup, T. Chem. Rev. 1995, 95, 1253.
(b) Fensterbank, L.; Malacria, M.; Sieburth, S. Synthesis
1997, 813. (c) Gauthier, D. R. Jr.; Zandi, K. S.; Shea, K. J.
Tetrahedron 1998, 54, 228.
(21) (a) Khan, A. R.; D’Souza, V. T. J. Org. Chem. 1994, 59,
7492. (b) Benazza, M.; Massoui, M.; Uzan, R.; Demailly, G.
Carbohydr. Res 1995, 275, 421. (c) Benazza, M.; Beaupere,
D.; Uzan, R.; Demailly, G. Carbohydr. Res. 1991, 218, 75.
(22) Djuric, S. W. J. Org. Chem. 1984, 49, 1311.
(23) Hodosi, G.; Podanyi, B.; Kuszmann, J. Carbohydr. Res.
1992, 230, 327.
(24) (a) Gorin, B. I.; Riopelle, R. J.; Thatcher, G. R. J.
Tetrahedron Lett. 1996, 37, 4647. (b) Baer, H. H.;
Berenguel, A. V.; Shu, Y. Y. Carbohydr. Res. 1992, 228,
307. (c) Gadelle, A.; Defaye, J. Angew. Chem. Int. Ed. Engl.
1991, 30, 78.
(25) Barluenga, J.; Campos, P. J.; Gonzalez-Nunez, E.; Asensio,
G. Synthesis 1985, 426.
(19) Cohen, Y.; Kotlyar, V.; Koeller, S.; Lellouche, J.-P. Synlett
2001, 1543.
(26) De Luca, L.; Giacomelli, G.; Porcheddu, A. Org. Lett. 2002,
4, 553.
(20) Chmurski, K.; Defaye, J. Tetrahedron Lett. 1997, 42, 7365.
Synlett 2004, No. 3, 564–571 © Thieme Stuttgart · New York