1
CHAPTER 1
1.1 Introduction
The most efficient organic reactions tolerate unprotected functional groups, generate
products in high enantiomeric excess and dramatically increase the molecular complexity
in a single step. The ideal situation arises when one can perform multiple bond
formation/fragmentation reactions sequentially in one pot. These may be carbon-carbon,
carbon-heteroatom or heteroatom-heteroatom bond manipulations. The difficulties
involved in designing these ideal reactions are considerable, with high selectivity and
high yield of each reaction being imperative. The efforts of good design are, however,
well rewarded by a successful reaction sequence due to the limited isolation and
purification steps that would otherwise be necessary. As the number of overall reactions
within a sequential reaction process increases, its complexity rises almost exponentially
and one finds that only a select group of reagents is able to face up to the considerable
task.
One such reagent capable of sequential processes is samarium diiodide (SmI2). 1 The
reagent was first discovered in the late 1970s and has dramatically changed the field of
organic synthesis. A large part of samarium diiodide’s potency stems from the fact that it
is capable of promoting both radical and anionic chemistry. Examples of these processes
include radical cyclisations, ketyl-olefin coupling reactions, pinacol coupling reactions,
Barbier-type reactions, aldol-type reactions, conjugate additions, and nucleophilic acyl
substitutions.
2
The versatility of SmI2 can be further enhanced by the addition of catalysts,1a by solvent
effects 2 or through other variations of other reaction conditions. 3 The ability to fine tune
the activity of the SmI2 allows it to be used under both sensitive and stringent conditions.
Depending on the type and rate of addition and any additives employed, samarium
diiodide can promote both one and two electron processes or a combination of the two.
These may be sequential radical processes, first radical and then anionic processes,
tandem anionic processes or carbanionic reactions followed by radical reactions.
1.2 Individual Reactions Promoted by Samarium(II) Iodide
1.2.1 Single-Electron Processes
There are a number of different functional groups that SmI2 can reduce to generate
radicals. Alkyl, aryl and alkenyl radicals can be produced from halide substrates while
ketyl radical anions are produced from aldehydes and ketones. The rate and selectivity of
the different reductions depend on the type of halide used and are often substrate
dependent as well.
1.2.1.1 Alkyl, Aryl and Alkenyl Radical Reactions
The initiation of the radical reactions usually takes place by the reduction of organic
halides. The intrinsic limiting factor in this process is that SmI2 can further reduce the
radical formed to the corresponding carbanion (Scheme 1). Thus, any desired radical
3
reaction (krad) must take place significantly faster than the reduction to the anion (kred).
This reactant is often employed instead of tin hydride and silicon hydride reagents. A
distinct advantage of this approach is the fact that an organosamarium species may be
formed if necessary for a sequential process. For a range of alkyl halides, reactivity of the
substrate decreases in the order: I>Br>Cl.1a
SmI2
R-X
SmI2
kredn
R-SmI2
E+
R-E
3
4
R
2
1
'R
SmI2
krad
5
(reaction)
'R-SmI2
E+
'R-E
6
7
Scheme 1
Radical processes promoted by SmI2 will only be synthetically useful if the radical has a
sufficient lifetime to undergo the desired radical process. Although many intramolecular
processes exist that meet this requirement, there is only a single class of intermolecular
process for which this rings true, namely the SmI2-promoted radical chain coupling of
polyhaloalkanes to alkenes and alkynes (Scheme 2). 4
F F
SmI2
+ Cl(CF2)2I
8
Cl
THF, 80%
9
10
Scheme 2
I
F F
4
The most common form of radical process is radical cyclisation. Of the more useful types
of reactions are the cyclisations of 6-halohex-1-ynes (Scheme 3). 5 The reduction of the
halo atom produces a non stabilised alkyl radical intermediate, which then cyclises onto
the alkynyl moiety to give the cyclopentylidene radical. This then abstracts an hydrogen
atom from the solvent or another suitable donor. It should be understood that radicals
such as aryl or alkenyl radicals abstract donor atoms from their surrounds faster than
SmI2 can reduce them. 6 This is an important consideration to take into account when
attempting a sequential process.
R
R
X
3 SmI2
R
R
SH
+S
THF, DMPU
12
11
13
14
15
Scheme 3
Substrates containing an oxygen atom within the chain are generally reduced more
rapidly than their all-carbon analogues. This phenomenon has been exploited in the
synthesis of lactones (Scheme 4). 7
n-Pr
Br
BnO
n-Pr
1. 2.5 SmI2
O
THF, HMPA, t-BuOH
2. H2CrO4
16
O
O
17
Scheme 4
5
Another useful type of alkyl radical can be derived from N-[(N’,N’-dialkylamino)alkenyl]benzoytriazoles (Scheme 5). Mixtures of diastereomers are usually
obtained in these reactions, with the cis isomers generally predominating. 8
O
O
SmI2
N
EtO2C
N
CO2Et
THF
N
N
18
19
N
Scheme 5
The formation of alkyl radicals is not limited to halogens, sulfones and related functional
groups, but can be formed by reductive cleavage of cyclopropyl ketones via a
rearrangement process (Scheme 6). The resultant radical is usually trapped
intramolecularly producing bicyclic systems. 9 These cleavage reactions will be dealt with
in more detail in a later section (see Chapter 1.5.2.2).
O
O
SmI2
TMS
THF, HMPA
20
TMS
21
Scheme 6
Although alkenes and alkynes are the most commonly used radical acceptors, some
workers have made use of hydrazones. Despite both SmI2 and tin hydride having been
6
utilised for the cyclisations of halohydrazones to cyclopentylhydrazines, 10 SmI2 is the
preferred reagent giving better diastereoselectivies at lower temperatures (Scheme 7).
X
NNPh2
H
NNPh2
4.5 SmI2
H
NNPh2
+
THF, HMPA
23
22a: X = Br
22b: X = I
7:1 ds
24
11: 1 ds
Scheme 7
The observed diastereoselectivity can be rationalised by viewing the two possible
transition state chair conformations. The structure leading to a trans product suffers from
an unfavourable 1,3-diaxial interaction between the substituent at the radical centre and
one of the hydrogen atoms on the ring, while this high energy species is absent in the
chair conformation of the cis product’s transition structure (Scheme 8).
H
N NPh2
H
R
25
Disfavoured
H
N NPh2
R
H
26
Favoured
Scheme 8
Benzofurans, naphthofurans and indoles have been synthesised using aryl radical
methodology.2d,6 The reactions proceed smoothly in the presence of tetramethylguanidine
(TMG), which seems to be a superior additive compared to HMPA or DMPU for these
types of reactions (Scheme 9).
7
Br
SmI2
N
N
THF, HMPA
28
27
Scheme 9
1.2.1.2 Ketyl Radical Anion Reactions
The ketyl radical-olefin coupling reaction is to date the most studied of all the different
processes that samarium(II) iodide promotes. The coupling serves as an highly useful
procedure to couple aldehydes and ketones to alkenes and alkynes.
The intermolecular version, although popular, has been used less frequently than its
intramolecular counterpart. This is due mainly to the fact that the reaction is restricted to
alkenes or alkynes that are good electrophiles. The products of these reactions are their
corresponding lactones (Scheme 10). 11
CHO
+
29
CO2Et
2 SmI2
THF, t-BuOH
30
O
O
31
Scheme 10
In addition to conjugated esters and nitriles, other conjugated alkenes such as styrene,
alkenylsilanes, vinyl acetates and allylic acetates serve as efficient ketyl radical acceptors
in SmI2-promoted intermolecular coupling reactions. 12 The reactions with unactivated
8
alkenes are, however, not as selective as the activated ones and often lead to mixtures of
isomers (Scheme 11).
O
+
Ph
OH
2 SmI2
OTMS
32
33
THF, HMPA
t-BuOH
Ph
O
34
Scheme 11
Alkynes also serve as suitable acceptors for ketyl radicals in intermolecular coupling
reactions (Scheme 12). 13 The carbon-carbon bond formation always takes place at the
terminus of external alkynes and at the most electropositive carbon atom of internal
alkynes. An activating group is preferential in obtaining suitable yields.
O
+
Ph
35
OH
2 SmI2
Ph
TMS
36
THF, HMPA
t-BuOH
Ph
TMS
37
Ph
Scheme 12
More commonly employed are the intramolecular ketyl olefin coupling reactions. These
are quite efficient and a wider range of acceptors is tolerated. This methodolgy has been
used to synthesise a number of ring structures varying in size from strained three
membered rings to larger eight membered ring systems.
Chelation can be used to control the relative stereochemistry of the hydroxyl and
carboxylate stereocentres formed during the cyclisation.6b Chelation of the two carbonyl
groups, by the Sm(III) Lewis acid generated, provides a template for stereochemical
9
control, providing a facial bias in the attack of the ketyl radical onto the unsaturated ester
(Scheme 13).
O
O
HO
OEt
EtO2C
CO2Et
2 SmI2
THF, t-BuOH
CO2Et
39
38
Scheme 13
The proximity of hydroxyl groups within the substrate has a significant influence on the
stereoselectivity of the reaction. 14 Even in the absence of additives such as HMPA, these
reactions proceed smoothly at low temperature with good yields and high
diasteroselectivity (Scheme 14). The addition of HMPA actually decreases the
stereoselectivity, thus demonstrating that the chelation of the Sm(III) species by the
ketone and hydroxyl groups plays a significant role.
OH
OH
SmI2
THF, MeOH
O
CO2Me
OH
CO2Me
41
40
Scheme 14
Although the carbonyl olefin coupling to form carbocycles is the most widely used
protocol, activated alkenes have been employed in the preparation of nitrogen
10
heterocycles (Scheme 15). 15 The yields of these reactions are, however, low and a
mixture of isomers is isolated.
O
N
MeO
OHC
O
OPh
OTBS
2 SmI2
N
THF, HMPA MeO
t-BuOH
HO
OPh
TBS
43
42
Scheme 15
The radical addition-elimination reaction of ketones with allyl sulfides has been
investigated in the synthesis of (-)-grayanotoxin III.16 The diasteroselectivities are quite
high although the observed olefin geometries are not stereospecific, when applicable. The
reaction can also be carried out with allyl sulfones (Scheme 16).
BnO
Y
3 SmI2
CHO
THF, HMPA
44a: Y = SPh
44b: Y = SO2Ph
OH
BnO
45
Scheme 16
The coupling reactions are not limited to alkenes and similar results have been achieved
using activated alkynes. Carbocycles, nitrogen heterocycles and oxygen heterocycles
have been prepared in this way (Scheme 17). 17
11
TMS
O
N
t-Boc
2 SmI2
TMS
HO
THF, HMPA
t-BuOH
N
t-Boc
47
46
Scheme 17
The ketyl radical is distinguished from its alkyl radical counterpart by the fact that it can
undergo cyclisation onto unactivated alkenes, alkynes and other functional groups. The 5exo-trig cyclisation represents one of the simplest of these transformations.
The
diastereoselectivity generally varies with the type of substituent next to the ketone
(Scheme 18). 18
O
2.2 SmI2
HO
R
THF, HMPA
t-BuOH
R
48a: R=Me
48b: R=Ph
HO
R
+
49
150:1
1:150
50
Scheme 18
The outcome of the reaction can be rationalised on the basis of a chair-like transition
structure (Scheme 19). The favoured state 51 predominates due to electronic effects as
long as the substituent R is small. As the steric bulk of R increases the disfavoured state
52 becomes favourable due to axial-axial interactions between R and the methylene
group in 51.
12
OSmI2-Ln
OSmI2-Ln
R
R
H
Disfavoured
Favoured
51
52
Scheme 19
The 5-exo-trig bicyclisation reactions promoted by SmI2 generally occur with high
diastereoselectivities. Both linearly fused bicyclics (Scheme 20)13 and bridged systems
(Scheme 21) have been synthesised using this methodology. 19
O
Et
2.2 SmI2
HO
THF, HMPA
t-BuOH
53
54
Scheme 20
O
2.2 SmI2
OH
H
THF, HMPA
t-BuOH
Me
56
55
Scheme 21
Similarly, alkynes have also been chosen to be radical acceptors (Scheme 22).17 Yields
are modest for unactivated alkynes but can be improved by activating them with
substituents such as trialkylsilyl groups. Although ketone ketyl precursors seem to
13
provide higher yields than their aldehyde counterparts, an aldehyde/alkyne coupling has
been used in the synthesis of isocarbacyclin. 20
HO
CHO
2.2 SmI2
C5H11
TBDMSO
THF, HMPA
t-BuOH
TBDMSO
C5H11
OH
OH
58
57
Scheme 22
The coupling of allenes is also possible, although only a limited number of reactions has
been studied (Scheme 23). 21
O
O
2 SmI2
MeO
O
MeO
+
CHO THF, t-BuOH
OH
60
59
MeO
OH
2:1
61
Scheme 23
Fewer studies of 6-exo cyclisation have been carried out as they suffer from lower yields
and selectivities (Scheme 24).18 The scope of these reactions is similar to those discussed
in the 5-exo-trig cyclisation case. They include cyclisation of the ketyl radical onto
alkenes, alkynes and allenes to form both cyclic and bicyclic structures.
14
O
2.2 SmI2
HO
THF, HMPA
t-BuOH
62
63 17:1 ds
Scheme 24
Pinacol coupling reactions of ketyl radicals have been well studied. The reactions of
aromatic compounds are highly substrate dependant as well as sensitive to the reaction
conditions and any additives employed. 22 The treatment of benzaldehyde with SmI2, for
example, leads to an almost quantitative yield of the corresponding pinacol product.
Upon repeating the reaction in the presence of HMPA, however, the aryl coupled product
is obtained in good yield with hydrobenzoin being the minor isomer (Scheme 25). This
observation is consistent with the notion that successful pinacol coupling reactions
require simultaneous complexation of both carbonyl moieties to one Sm ion.
CHO
SmI2
OH
+
CHO
HO
OH
66
65
64
THF
0%
100%
THF/HMPA
60%
10%
Scheme 25
The coupling of non-aromatic substrates is much simpler. In the absence of a proton
source such as methanol, aldehydes and ketones are easily coupled in the presence of
SmI2 to their corresponding pinacol products (Scheme 26). 23 If a proton source is indeed
15
present, the carbonyl group undergoes simple reduction to give the respective alcohols.
n-C6H13C(O)CH3
OH
SmI2
THF
67
n-C 6H13
n-C 6H13
OH
68
Scheme 26
In the same family of reactions is the coupling of aldimines using SmI2 (Scheme 27). 24
Both alkyl and aryl aldimines can be coupled.
SmI2
PhCH NPh
THF
69
NHPh
Ph
Ph
NHPh 70
Scheme 27
Cyclisation reactions of 1,5- and 1,6-dialdehydes or diketones using the pinacol
methodology proceed smoothly with considerable stereochemical control: the cis-diols
are almost exclusively obtained (Scheme 28). 25 Heteroatom functionality α to the
carbonyl group is tolerated showing that pinacolisation is much faster than reductive
cleavage. Furthermore, polar substituents α to the carbonyls end up anti to the diol
stereocentres in the final product.
OBn
TBDMSO
CHO
O
OBn
SmI2
OH
THF
OH
TBDMSO
71
72
Scheme 28
16
Although ketyl radical/nitrile coupling is possible, the coupling process is slow and the
yield is low (Scheme 29). This is indicative of the reluctance of the nitrile moiety to
undergo radical addition.6b,26
O
O
OH CO2Et
SmI2
OEt
THF, HMPA
NC
73
O
74
Scheme 29
A somewhat better radical acceptor is the hydrazone moiety. Cyclisation of carbonylhydrazones provides an excellent route to 2-aminocyclopentanols (Scheme 30).10
N
NPh2
H
O
n
NHNPh2
SmI2
R
R
THF, HMPA
n
75
R = H,Me
n = 1,2
OH
76
Scheme 30
1.2.2 Two-Electron Processes
The ability of samarium diiodide to promote both one- and two-electron processes sets it
apart from most other reductive coupling reagents presently available. The two-electron
process complements the more common organolithium, organomagnesium and
organozinc chemistry.
17
1.2.2.1 Barbier- and Grignard-Type Reactions
The samarium Grignard reaction is limited in scope to primary and secondary halides and
even then the reaction does not match up to its organolithium and organomagnesium
counterparts.
The samarium Barbier reaction is just the opposite, succeeding where the magnesiumand lithium-promoted reactions fail. Samarium diiodide promoted reactions also have a
broader scope and tolerate a wider range of functionalities. Although the intermolecular
Barbier reaction is limited in use, the intramolecular form is widely used and has found
an important niche in organic synthesis.
Allylic, propargylic and benzylic halides are highly reactive in samarium Barbier
reactions, with both aldehydes and ketones serving as partners for the reactions (Scheme
31).26
HO
CHO
I
+
2 SmI2
THF
77
78
79
Scheme 31
The active functionality can be extended to sulfones. These reactions require HMPA as
an additive to ensure a successful reaction (Scheme 32). 27
18
SO2Ph + n-C 6H13CHO
OH
4 SmI2
n-C 6H13
THF, HMPA
81
80
82
Scheme 32
Selective transformations of polyfunctional substrates can be accomplished by taking
advantage of the difference in reduction potentials between alkyl iodides, bromides and
chlorides (Scheme 33).1a
2 SmI2
I(CH2)6Cl
+
n-C 6H13C(O)CH3
83
THF
HO
n-C 6H13
84
(CH2)6Cl
85
Scheme 33
The most promising of the samarium Barbier reactions are the intramolecular processes
that form five or six membered rings upon cyclisation. These reactions allow easy access
to cyclopentanol 28 and cyclohexanol 29 derivatives (Scheme 34).
O
Br
2 SmI2
OH
THF
86
87
Scheme 34
The same protocol has been applied to diverse systems and in virtually every case has
been superior to conventional methods. Suginome and Yamada successfully employed
this technique in the synthesis of (±)-exaltone and (±)-muscone (Scheme 35). 30
19
O
HO
2 SmI2
I
H
THF, HMPA
89
88
Scheme 35
Another use of the reaction is that of making bridged bicyclic alcohols (Scheme 36). 31
The methodology has been applied to make highly strained and complex molecules that
are otherwise accessible only with some difficulty.
O
OH
2 SmI2
THF, cat. Fe(III)
I
90
91
Scheme 36
1.2.2.2 Aldol- and Reformatsky-Type Reactions
The Reformatsky reaction promoted by samarium(II) iodide has been carried out between
α-halo esters and ketone electrophiles (Scheme 37).1a, 32
O
CO2Et
2 SmI2
HO
CO2Et
+
Br
92
THF, HMPA
93
94
Scheme 37
20
Similarly aldol-type reactions between α-halo ketones and suitable aldehydes have
proven successful (Scheme 38). 33
O
Br
+
O2N
CHO
THF, HMPA
95
O
2 SmI2
96
Ph
NO2
97
Scheme 38
Condensations between α-keto carboxylates and α-halo ketones have also been attempted
with good yield. 34
The intramolecular reaction, as before, finds more use in organic synthesis. Reactions of
SmI2 with β-bromoacetoxy carbonyl substrates are proposed to generate a Sm(III) ester
enolate with cyclisation taking place through a rigid cyclic transition structure enforced
by chelation (Scheme 39). 35
O
Br
H
O
O
SmI2
R1
R3
R2 H
98
R3
R2
H
R1
R2
R1
OH
O
O
99
Scheme 39
O
Sm(III)
R3
O
100
O
21
1.2.2.3 Nucleophilic Acyl Substitution Reactions
Although unreactive in intermolecular reactions, esters and amides undergo highly
selective nucleophilic acyl substitution reactions when appropriately substituted
carboxylic acid derivatives are treated with SmI2 (Scheme 40). 36 The method is amenable
to the synthesis of four-, five- and six-membered rings with primary alkyl, secondary
alkyl and allylic halides participating in the reaction.
O
O
I
2 SmI2
Y
Ph
Ph
THF
101a: Y = OEt
102
101b: = S(i-Pr)
101c: = N(OMe)Me
71%
68%
81%
Scheme 40
1.3 Sequential Reactions Promoted by Samarium(II) Iodide
The ability of SmI2 to promote sequential processes has been a major factor contributing
to its popularity as a reducing agent in organic synthesis. Another extremely useful
property of the reagent is that the individual reactions within a reaction sequence may be
either one- or two-electron processes. As the sequential processes are just a combination
of the above mentioned individual chemistries, only a limited number of examples shall
be covered for each set.
22
1.3.1 Sequential Radical Processes
The success of tin hydrides and silicon hydrides in promoting sequential radical processes
has overshadowed SmI2 as a potent reagent for the above and thus very few examples of
SmI2-mediated sequential radical processes have been reported. One such example has
been employed in the key step in the synthesis of (±)-coriolin (Scheme 41). 37
CHO
O
1. 1.3 SmI2, HMPA
O
2. p-TSA, acetone
HO H H
HO H H
O
H
103
91%
O
+
+
H
104
105
OH
O
O
106
Scheme 41
1.3.2 Radical/Anionic Sequences
These are the most widely studied and most commonly used sequential processes.
Generally, a radical intermediate is formed after SmI2 has reduced a suitable precursor.
The radical intermediate then undergoes an addition onto a radical acceptor forming a
second radical intermediate. This is then further reduced by a second equivalent of SmI2
to form the organosamarium that can be trapped by a number of electrophiles (Scheme
42). 38 Only primary or secondary radicals can be reduced to the organosamarium with
tertiary radicals abstracting an hydrogen atom or undergoing disproportionation.38
23
I
1. 2 SmI2
O
O
2. 3-pentanone
OH
108
107
Scheme 42
1.3.3 Anionic/Radical Sequences
The anionic/radical process is characterised by the addition of four or more equivalents of
SmI2. The first two equivalents are responsible for producing the organosamarium that
normally undergoes a nucleophilic acyl substitution reaction. The resulting ketone is then
reduced by one equivalent of the remaining SmI2 and further undergoes a radical
cyclisation process. Subsequent reduction of the radical and protonation yields the
desired multistep product (Scheme 43).36,39 The sequence can be repeated by quenching
with a suitable electrophile instead of a proton source.
O
I
OEt
4 SmI2
O
HO
THF, HMPA
109
110
111
Scheme 43
1.3.4 Sequential Anionic Processes
The most straightforward reaction of this type is one in which two intramolecular Barbier
type reactions occur in one pot. Although Molander has classed this particular reaction as
24
sequential, it is clear that two separate reactions occur at two different centers. This
strategy has been used in the synthesis of polyquinenes (Scheme 44). 40
Br
H
4 SmI2
O
O
THF, HMPA
OH
HO
H
Br
113
112
Scheme 44
1.4 HMPA - Effects and Mechanistics
The fact that Sm(II) is multivalent suggests that the redox potential of a divalent Sm
species will vary depending on the number and type of ligands coordinated to it.
Although HMPA 41 is the most widely used co-solvent in SmI2 mediated reactions, other
co-solvents containing basic oxygen including 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)pyrimidinone (DMPU) 42 and 1,1,3,3-tetramethylurea (TMU) 43 have been used with
success.
1.4.1 The Effect of HMPA on the Reduction Potential of SmI2
Although there were numerous synthetic examples of the ability of HMPA to enhance the
yield and selectivity of a SmI2 mediated reaction,1 it was not until the work of Flowers
that this effect was quantitatively studied. 44
25
The change in redox potential of a reducing agent may significantly alter the mechanistic
pathway that a reaction follows, possibly with deleterious effects. The determination of
the redox potentials of a series of Sm(II) species is thus of utmost importance when
designing a reaction sequence. Species that do not succumb to SmI2 in THF may be
readily reduced within a SmI2-HMPA system.
One way in which to determine the energetics of an one-electron reducing system is to
make use of electrochemistry. Flowers studied the effect of HMPA by adding aliquots of
HMPA to a SmI2-in-THF system and subsequently recording a linear sweep
voltammogram for each co-solvent addition (Table 1). Only a slight change in the
reduction potential is observed after the addition of up to two equivalents of HMPA.
Entry
Equivalent of
Oxidation
ΔE, V (kcal)
HMPA vs. SmI2a
Potential Vb
1
0
-1.33
0
2
1
-1.43
0.10 (2.3)
3
2
-1.46
0.13 (3.0)
4
3
-1.95
0.62 (14.0)
5
4
-2.05
0.72 (16.6)
6
5
-2.05
0.72 (16.6)
7
6
-2.05
0.72 (16.6)
a) conc. of SmI2 = 0.5 mM. B) vs. Ag/AgNO3 reference electrode in THF.
Table 1
26
However, on addition of the third equivalent, one sees a more drastic increase in the
reducing power of the SmI2-HMPA species from –1.46 V to –1.95 V. On addition of the
fourth portion, one again sees a significant (although less so) increase but this plateaus on
the addition of five or more equivalents.
Hou and co-workers 45 have successfully isolated and obtained crystallographic data of a
SmI2(HMPA)4 complex which they believed to be the reactive species in the SmI2HMPA mediated reaction. This is consistent with the above electrochemical
investigation. Another validation of these results comes from the work of Curran. 46 He
investigated the rate constants of the reactions of primary alkyl radicals promoted by
SmI2 with varying HMPA concentrations, and observed that the reaction time of 6-iodo1-hexene was too slow for their rate experiments in the absence of HMPA. Upon addition
of 2 equivalents of HMPA the rate became measurable and reached a maximum on
addition of the fifth equivalent.
1.4.2 A Structural Basis for the HMPA Effects
HMPA is not only responsible for increasing the reduction potential of SmI2 but also
influences the regio- and stereo-selectivity of SmI2-promoted reactions. 47 Despite the
extensive use of the SmI2-HMPA complex in organic synthesis, the exact role played by
HMPA in these reactions was not well understood prior to the isolation and structure
determination of the SmI2-HMPA complex by Hou and co-workers. 48 They were posed
with the following questions: a) Why does HMPA increase SmI2’s reducing ability? b)
Why is the regio- and stereochemistry of the reaction altered upon addition of HMPA? c)
27
How much HMPA is necessary for optimum results? d) Why does the selectivity become
poorer when the HMPA exceeds a certain amount? To get a better understanding of the
concepts involved in these effects it was necessary to isolate and structurally characterise
various SmI2-HMPA complexes.
[SmI2-(HMPA)4] The single crystal was obtained by adding 4.5 equivalents of HMPA to
a solution of SmI2 in THF, and subsequently crystallising it from toluene (HMPA binds
to the metal through the oxygen atom on the phosporus). The structure reflected a slightly
distorted octahedral complex (Figure 1), the most notable feature of which was the short
Sm-O(HMPA) bonds (av. 2.500 Å) whereas other oxygen-donor ligands have longer SmO bond distances (av. 2.600 Å). 49 Due to the strong co-ordination of the four HMPA
ligands the Sm-I bonds were also much longer (av. 3.390 Å) than those found in other
SmI2 complexes such as cis-[SmI2(dme)(thf)3] (av. 3.246 Å). This was indicative of
weaker Sm-I bonds. The four HMPA ligands as well as the iodides were bound in the
innersphere.
SmI2 + HMPA
1
:
4.5
THF
I
HMPA
HMPA
Sm
HMPA
HMPA
I
114
Figure 1
SmI2-(HMPA)6] The complex was isolated after the addition of 10 equivalents of HMPA
to a solution of SmI2 in THF. The excess was added in order to establish the maximum
28
number of ligands that would co-ordinate to the Sm(II) ion. The complex was also
octahedral in shape but featured the iodide ions in the outer sphere position (Figure 2).
Due to the six bulky HMPA ligands, the Sm-O bonds in this complex were slightly
longer than before (av. 2.531 Å) but still considerably shorter than those between Sm(II)
and other oxygen-donor ligands.
2+
SmI2 + HMPA
1
:
HMPA
THF
HMPA
10
HMPA
HMPA
Sm
HMPA
HMPA
-
2I
115
Figure 2
Based on the information presented by the crystal structures, it becomes clear that
complexation of HMPA to the Sm(II) ion greatly changes both the electronic and steric
environments around the metal. The long Sm-I bond distance suggests that the bond is
relatively weak and is easily cleaved by another co-ordinative ligand. The Sm(II) metal
centre will accept as many as six strongly electron-donating ligands when a sufficient
amount of HMPA is added. This would explain the enhancement of the reduction
potential of SmI2 upon addition of HMPA.
When four equivalents of HMPA are added a square planar co-ordination geometry of the
ligands around the metal ion is adopted, which has been used to explain the selectivity
observed in many SmI2-mediated reactions. It is proposed that a substrate will approach
the metal centre from either above or below the square plane formed by the HMPA
29
ligands. The steric repulsion between the substrate and the bulky ligands will naturally
change the regio- and stereoselectivity of the reactions as compared to those carried out
in the absence of HMPA.47
When six or more equivalents of the co-ordinating ligand are used, the Sm-I bonds are
cleaved and the metal centre is surrounded by six bulky HMPA ligands. In this scenario
the approach of substrates is blocked and the electron-transfer reaction takes place in the
outer sphere.
The study shows that if one is interested only in increasing the reducing power of SmI2,
the use of six or more equivalents of HMPA is appropriate. If, however, one would like to
control the selectivity of a reaction, only a limited amount of the ligand should be added.
Although most information concerning the structure of both the Sm(II) and Sm(III)
oxidation states has been provided through crystallographic studies, little work has been
done on the structure of complexes of Sm and HMPA in solution. Flowers and coworkers
have studied the structure and energetics of the samarium diiodide-HMPA complex in
THF using UV-vis spectroscopy, isothermal titration calorimetry and vapor pressure
osmometry. 50 . They found that the aggregation number for SmI2 in THF was 0.98 ± 0.09
over the entire concentration range studied, indicating that SmI2 is monomeric as well as
the fact that the addition of HMPA cosolvent to a solution of SmI2, displaces the THF
bound to the metal. If four equivalents of HMPA are added the iodide ions remain
30
innersphere while on the addition of six or more equivalents of HMPA forces the iodide
ions to an outersphere position.
Crystal structures of Sm(III) complexes indicate an octahedral environment about the
metal centre.
1.5 Carbon-Carbon Bond Fragmentation Reactions
1.5.1 C-C fragmentation reactions in ‘no-strain’ systems
Although various dehalogenation and deoxygenation reactions of α-hydroxy, α-alkoxy, or
α-acyloxycarbonyl compounds utilising samarium(II) iodide have been extensively
investigated, comparatively little work has been done on C-C bond fragmentation
reactions.
Magnus and co-workers published one of the first C-C bond fragmentation reactions
promoted by SmI2 (Scheme 45). 51 The reagent was resorted to after a reaction with
Bu3SnH in toluene (40 h at reflux) yielded a mixture of the desired fragmentation product
and the product of simple reduction in modest yield. Upon subjecting the steroid
derivative to SmI2 in THF at room temperature, the C-C cleaved product was isolated in
88% yield in only five minutes, and none of the reduced product was detected. The
authors suggested that a plausible explanation for the success of the reaction using SmI2
where the tin reagent failed was the ability of the Sm(II) to rapidly reduce a radical to an
anion. The radical mechanism can, however, not be discarded.
31
O
MeS
S
O
SmI2
H
H
O
H
O
H
H
H
+
THF
H
Me
O
H
O
HO
116
117
118
88%
0%
Scheme 45
Honda and co-workers 52 have successfully prepared physiologically active, chiral, natural
products using a carbon-carbon bond cleavage reaction as the key step. Starting from a
chiral cyclopentane derivative, they were able to regioselectively cleave the α,β-C-C
bond with respect to the ester functionality (Scheme 46). Since SmI2 can complex both
the halogen atom as well as the carbonyl moiety γ to the Cl, forming a seven-membered
ring transition state, a concerted fragmentation mechanism was proposed in converting
the carbocycle into the acyclic alkene. The reaction was successful using both the bromo
and chloro substrates.
OTES
OTES
3 SmI2
Cl
CO2Me
CO2Me
THF, HMPA
86 %
119
-
OTES
OMe
Cl
O
[Sm2+]
121
Transition State
Scheme 46
120
32
In the absence of HMPA, none of the fragmented product was isolated: reductive
dehalogenation took place in good yield. The reactions were proved to be SmI2 specific
with no fragmentation taking place under other standard reducing conditions. These
included the treatment of the ester with tri-n-butyltin hydride and azobisisobutyronitrile
(AIBN) in refluxing benzene under radical-initiating reaction conditions, and the
reduction of the substrate with zinc powder in acetic acid.
The authors were cautious in proposing a mechanism for the observed fragmentation.
However, they speculated that the results suggest that the reaction would occur via a twoelectron reduction process, involving further reduction of the initially formed radical
anion prior to either proton abstraction or a radical-induced fragmentation (see transition
state above).
Reaction of the ketone analogue of the ester, i.e. the γ-halo ketone, also yielded the bond
cleaved product. In contrast, treatment of open-chain γ-halo ketones (halogen = Cl, Br)
with SmI2 under the same conditions has been known to yield cyclobutanols, via the ketyl
radical. 53 The structural bias or inherent strain in forming a bicyclo[3.2.0]heptane ring
system is probably the deciding factor which promotes the fragmentation over
cyclisation.
Honda has used the SmI2 promoted fragmentation of γ-halo esters for the synthesis of a
number of enantiomerically pure natural compounds including alkaloids (Scheme 47),
terpenes and antibiotics. 54
33
OTBDMS
Cl
CO2Me
O
122
N
H
123
Scheme 47
The fragmentation reaction of an ε-halo-α,β-unsaturated ester, where a similar cleavage
reaction involving the fragmentation of the carbon-carbon bond between the γ and δ
positions of the carbonyl group, was also investigated (Scheme 48). 55 The substrate
underwent bond scission at room temperature upon treatment with SmI2 / HMPA to give
a 2:1 ratio of stereoisomers (E:Z), in 91% yield.
OTES
OTES
SmI2
O
O
THF, HMPA
Cl
O
CO2Et
CO2Et
125
124
OH
OH
Nemorensic acid
126
Scheme 48
Molander has described stereocontrolled cyclisation reactions of a number of substrates
mediated by SmI2. 56 Intramolecular cyclisation of allylic halides of varying chain lengths
onto
β-keto
esters
successfully
provided
vinyl-substituted
cyclopentanes
and
cyclohexanes in good yields as a mixture of isomers. However, the attempted cyclisation
of ethyl 2-methyl-2-(trans-4-bromo-2-butenyl)-3-oxobutanoate produced ethyl 2-methyl3-oxobutanoate instead of the desired cyclobutane or cyclohexane derivative. The authors
34
suggested that loss of butadiene as required for this transformation is facilitated by the
ability of a β-keto ester stabilised anion or radical intermediate to serve as an effective
leaving group in the reaction (Scheme 49). This result implied that the rate of
fragmentation is substantially faster than that of cyclisation, which would normally first
require reduction of the radical intermediate to the corresponding anion.
O
O
Me
O
OEt
Me
Br
2 SmI2
O
OEt
Me
Me
THF, - 78 °C
128
127
Scheme 49
This theory is supported by the outcome of the reaction shown below (Scheme 50). In
this case, fragmentation would result in a less stabilised anion/radical species than that of
the previously mentioned reaction; treatment of trans-8-bromo-4-methyl-6-octen-3-one
with SmI2 afforded 1-ethyl-6-methyl-3-cyclohexen-1-ol in 91% isolated yield.
O
Me
Me
Br
2 SmI2
Et
OH
Me
THF, - 78 °C
129
130
Scheme 50
Samarium(II) iodide has been successfully used to promote reductive decyanation of
malonitrile derivatives (Scheme 51). 57 The decyanation was achieved using a range of
35
either monosubstituted (R1 = alkyl, R2 = H) or disubstituted (R1 = alkyl, R2 = alkyl)
malonitriles in good yield. It was found that HMPA was essential for the success of the
reaction. In the substrates where one of the alkyl groups contained C-C unsaturation, no
cyclisation occurred where either a five- or seven membered ring could have formed. The
authors expanded on this work and were able to apply this new methodology to decyanate
α-alkoxycarbonyl substituted nitrile derivatives. Again, the addition of HMPA was
critical for the reaction pathway to proceed. Although Bu3SnH can be used for the
decyanation of malonitrile derivatives, the decyanation of the α-alkoxycarbonyl nitrile
compounds seemed to be SmI2 specific.
R2
CN
SmI2, THF
R2
CN
R1
CN
HMPA, 0 °C
R1
H
131
132
R2
CO2Et
SmI2, THF
R2
CO2Et
R1
CN
HMPA, rt
R1
H
134
133
Scheme 51
Although several methods of producing macrocyclic lactams have been forthcoming, the
construction of these large ring systems still presents a significant challenge because ring
closure is difficult to achieve. 58 With the discovery that alkylazides are reactive towards
reduction by SmI2 came an investigation into the ring enlargement of readily available
azidocyclododecanones to large-ring lactams (Scheme 52). 59 Although the exact
mechanism of the reaction is unknown, the synthetic route is mild yet efficient in
36
producing 16- and 17-membered lactams. Attempts to synthesize larger ring systems by
this route failed.
O
CN N
3
n
O
O
H
N
CN
SmI2, THF
NH
NH
+
rt, 2 h, 92%
(2:1)
n=3
135
136
137
Scheme 52
1.5.2 C-C fragmentation reactions in ‘ring-strained’ systems
1.5.2.1 Cyclobutane containing substrates
A more common form of C-C cleavage is that of strained ring systems, including
compounds that contain a cyclobutane moiety. The initial fragmentation of the following
cyclobutane substrate led to an allylic radical, which upon further reduction and
protonation yielded an isomeric mixture of fragmented products (Scheme 53). 60 The
exact ratio of the mixture was shown to be dependent on the nature of the reducing agent
employed. The reagents used were n-Bu3SnH, (C6H5)3SnH and SmI2, with AIBN used as
the initiator for the former two reagents.
37
CH2I
2 SmI2
+
93 %
R
99
R = H, Me, CO2R'
R' = Me, t-Bu
139
138
R
1
R
140
Scheme 53
With the hydride reagents the exocyclic double bond isomer predominated while with
SmI2 the endocyclic double bond product was favoured. It was established that activation
of the double bond was not necessary for the success of the reaction as displayed by the
two cases where R = H and R = Me. SmI2 was determined to be the reagent of choice in
these transformations with the fragmentation yields being in excess of 90 % in all four
cases.
The reactions were assumed to proceed via a radical pathway forming the allylic radical
after fragmentation of the strained cyclobutylcarbinyl system (Scheme 54). This radical
was then reduced to the carbanion by a second equivalent of SmI2, and finally protonated.
Support of this theory is provided by the fact that quenching with excess MeI gave only
the α-methylated product when making use of the ester substrates, which, at the very
least, proves that the reaction ends in a carbanion.
38
CH2I
SmI2
SmI2
CO2Me
CO2Me
141
142
CO2Me
143
MeI
Me
CO2Me
144
Scheme 54
Fragmentation of appropriate [2+2] photoadduct derivatives led to bicyclo[m.n.0]carbon
skeletons that are present in a wide range of natural products. This methodology has been
used in the synthesis of Dictamnol, a trinor-guaiane (Scheme 55). 61 The key step involves
the initial reduction of a diiodo compound and a subsequent free radical fragmentation.
Reduction of one iodo moiety initiates the reaction while the other iodide serves as a
leaving group in the last stage of the fragmentation sequence. Treatment of the substrate
with SmI2 in THF and DMPU provided the ring expanded 5,7 fused ring diene in good
yield. Apart from the generally higher yield offered by SmI2, this reagent is preferable to
n-Bu3SnH as the radical initiator because of the convenience, lower toxicity, and the ease
of product purification when using the former.
39
H
H CH2I
2 SmI2, THF
Me
OH H H H I
DMPU Me
72 %
OH H
146
145
Scheme 55
A very similar methodology has been used to synthesise the 5,7 ring system and the
strained cyclopropane moiety of the aromadendrane family of sesquiterpenoids (Scheme
56). 62 Again, the driving force for the reaction is the cleavage of the cyclobutane C-C
bond of the [2+2] photoadduct.
H CH2I
SmI2, THF
H H
DMPU, rt, 51%
CO2Et
H
147
H
CO2Et
CO2Et
149
Major product
148
+
H
H
EtO2C
150
Scheme 56
H
CO2Et
151
40
After fragmentation, the radical is trapped by a pendant α,β-unsaturated ester moiety,
leading to the formation of a cyclopropane ring. The stabilised radical (or anion?)
intermediate allows this apparently contra-thermodynamic process to take place.
Although the cyclopropane was the major product isolated, a small amount of an isomeric
mixture of α,β-unsaturated compounds was formed by cyclisation followed by opening
on either side of the cyclopropyl carbinyl system.
The examples discussed above describe the synthesis of a number of terpenoids in which
the critical step is the fragmentation of the “internal” cyclobutane bond. The cleavage of
the “external” cyclobutane bond, however, allows the formation of a key intermediate in
the preparation of the sesquiterpenoid trichodiene (Scheme 57). 63 In the critical step of
the synthesis, SmI2 facilitated the desired cleavage of the external cyclobutane bond, to
give the fragmented product in 95% yield. This fragmentation not only formed the
cyclopentylcyclohexane system needed, but also introduced the exocyclic methylene
group, which is a structural feature found in the natural product.
SmI2, THF
DMPU
I
152
O
95%
O
153
154
Trichodiene
Scheme 57
A number of physiologically active alkaloids contain a spirocyclic skeleton. This moiety
is easily accessed by a novel cyclobutane ring cleavage (Scheme 58). 64 The
41
regioselective ring opening was effected after treatment of the tricyclic precursor with
SmI2 in THF-DMPU, to give the desired spirocyclic ketone in 68% yield. The unique
biological activity (e.g. compounds of this structure have been used in probing the
mechanisms involved in transsynaptic transmission of neuromuscular impulses) of these
alkaloids has stimulated considerable interest in their synthesis.
O
O
H
H
R
N
Boc
155
SmI2, THF
R
DMPU
68%
R = α/β-n-C5H11
N
Boc
156
Scheme 58
1.5.2.2 Cyclopropane containing substrates
Alkyl radical cyclisations and tandem cyclisations are powerful aspects of the synthetic
chemist’s arsenal. The corresponding ring-opening fragmentation reactions, which in
most cases are disfavoured from both a kinetic and thermodynamic viewpoint, are,
however, scarcer. An exception to this rule of thumb is the cyclopropylcarbinylhomoallyl radical rearrangement, which is both fast and thermodynamically favoured as a
direct consequence of the cyclopropyl ring strain, as is neatly shown by the work of
Motherwell. 65 The beauty of the reaction is that ring opening occurs under
stereoelectronic control, leading to the fragmentation of the exocyclic C-C bond. The
regio- and stereochemistry of the intermediate is thus determined by the initially
constructed cyclopropyl ketone.
42
If no radical trap is incorporated into the molecule or added to the reaction mixture, only
the fragmented methyl substituted derivative is formed. If a radical acceptor is present,
the radical cascade reaction will dominate and allows entry into either spiroketones
(Scheme 59) or fused bicyclic systems depending on the connective placement of the
radical accepting chain. The reaction proceeds with alkenes and alkynes, and activating
electron-withdrawing groups enhance the yield of the reaction.
O
i) SmI2, THF
DMPU
OAc
ii) AcCl
157
57 %
158
Scheme 59
The reaction sequence, mechanism and intermediates allow not only tandem radical
cyclisation reactions, but also allow capitalisation of the enolate anion chemistry through
trapping of the intermediate samarium enolates with electrophiles in cases where
carbonyl-type alkene activating groups are present.
Attempted trapping with allyl bromide gave the allylated product in 37% yield and the
epimeric ketone products of simple ring fragmentation in a combined 9% yield (Scheme
60). These experiments showed that it is possible to trap the Sm enolates with allyl
bromide after a radical reaction sequence had taken place, in somewhat diminished
yields.
43
O
O
O
i) SmI2, THF
DMPU
+
ii) allyl bromide
159
160
161
37%
9%
Scheme 60
Although other reagents such as Bu3SnH or sodium naphthalenide can be used for this
type of radical chemistry, the SmI2/DMPU system was by far the most useful in terms of
avoiding the problems of reagent basicity and/or second electron transfer associated with
these reducing agents.
Subsequent to his first example of a simple cyclopropane ring cleavage (Scheme 61),52
Molander has further developed the cyclopropyl ring cleavage reaction by taking
advantage of the reducing strength of SmI2 for further transformations. 66
O
O
2 SmI2, cat. Fe(DBM)3
Me
THF, 69%
162
Me
163
Me
Scheme 61
The initially formed methylene radical can be further reduced to the corresponding
carbanion and be trapped intramolecularly by a number of electrophiles (Scheme 62).
These include ketones, esters, epoxides and aldehydes and lead to the formation of a
variety of functionalised spirocyclic, bicyclic and tricyclic ring systems.
44
O
O
2.5 SmI2
Me
n
m
8 HMPA
OH
n
m
O
164
Me
165a: n=0, m=1 61%
165b: n=1, m=1 79%
165c: n=1, m=2 30%
Scheme 62
The reductive ring opening of α-cyclopropyl ketones with SmI2 has been used in the key
step for the preparation of angular and linearly fused triquinanes (Scheme 63 and Scheme
64). The ring cleavages were effected by a SmI2-THF-MeOH system in good yield. 67
H
H
OMEM
O
OMEM
SmI2, THF, 25°C
MeOH, 86 %
O
H
H
167
166
Scheme 63
H
SmI2, -78°C to 25°C
BzO
O
H
BzO
THF, MeOH, 65%
O
H
OBz
OBz
168
169
Scheme 64
The samarium(II) iodide promoted ring opening of cyclopropylogous α-hydroxy carbonyl
compounds has been investigated as a possible strategy in the ongoing search for new
45
syntheses of the tricyclo[5.3.1.01,7]undecane system of taxanes. 68 Two types of substrates
were used in the study: a range of cis-substituted cyclopropanes (Scheme 65) and
compounds containing a bicyclo[3.1.0]system ring (Scheme 66). The reactions with
substrates containing aldehyde moieties were carried out in THF at room temperature in
the absence of HMPA, while for reactions with ketones, the addition of 8-10 equivalents
of HMPA was necessary. The former set of substrates formed a range of fragmentation
products including homoallylic ketones, δ-hydroxy ketones and β-methyl-γ-hydroxy
ketones in various amounts depending on the R group. The mechanism of the reaction
was proposed to proceed via the ketyl radical. This radical then undergoes one of two
different rearrangements depending on which C-C bond of the cyclopropane ring cleaves.
Further reduction and, in one case, a sequential β-elimination afforded the three different
fragmentation compounds.
O
OH
R
rt
170
O
O
SmI2, HMPA
+
R
171
O
R
R = alkyl
OH
172
+
Me
OH
R
173
Scheme 65
The results of the reactions of the bicyclic compounds with SmI2 were of greater interest.
The regioselectivity of the reaction depended on the nature of the carbonyl group present:
aldehydes underwent an endo bond cleavage followed by a β-elimination of the hydroxy
group to give cyclohexenes, whereas ketones formed their respective cyclopentanols
resulting from simple exo C-C bond fragmentation (Scheme 66). The rationalisation of
46
these findings relied either on a steric repulsion argument or one based on Frontier
Molecular Orbital theory.
OH
OH
SmI2
SmI2
R
(R = alkyl)
R
(R = H)
H
O
O
O
174
175
176
Scheme 66
The reductions of both α-haloketones to the corresponding enolates and the ring opening
of α-cyclopropyl ketones have both been investigated independently. Beerli et al. have
studied the effect of having both these functionalities present under reducing conditions
(Scheme 67 and Scheme 68). 69 Although a ring opening and elimination sequence took
place with a variety of reducing agents, only SmI2 and NaHTe gave good stereochemical
control. With systems such as Li/NH3, Zn/AcOH, Cr(III), and Bu3SnH/AIBN, mixtures
of the subsequent keto-alkene isomers were obtained, with the thermodynamically
favoured trans product dominating.
The fact that the reduction with SmI2 is highly stereoselective (selective production of the
cis or trans product, depending on the stereochemistry at the halogenated carbon atom),
in contrast with other reducing agents like Bu3SnH, led the authors to suggest a concerted
reaction pathway, which would be similar to a Grob fragmentation. 70 Whether the
reaction starts at the halogen or at the ketone is unclear, and it is possible that the
mechanism resembles that proposed earlier for the ring scission of γ-haloesters (see
section 3.1).38
47
Br
H
O
O
O
SmI2, THF
H
+
MeOH, rt
H
H
H
178
179
99
1
76 %
177
Scheme 67
O
Br
H
O
O
H
SmI2, THF
H
MeOH, rt
85 %
+
180
H
H
181
98
182
2
Scheme 68
The reduction and subsequent ring opening of α-halo-oxirane rings has been well studied.
These reactions are generally selective for carbon-oxygen bond cleavage, although
carbon-carbon bond fragmentation does occur when the oxirane ring possesses a vinyl or
aryl substituent, which would lead to a resonance stabilised carbon radical (Scheme
69). 71 Treatment of a bromomethyl epoxide with 2.2 equivalents of SmI2/THF in the
presence of HMPA and MeOH at ambient temperature afforded the analogous allyl
alcohol (1-phenyl-2-(1-naphthylmethoxy)ethane) as well as the vinyl ether (1-(1naphthyl)-3-phenylprop-2-enol) as a by-product.
Br
OH
SmI2, HMPA
Ph
O Nap
183
MeOH, THF
(Nap: 1-naphthyl)
Ph
Nap
184
Scheme 69
+ Ph
O
185
Nap
48
The presence of the vinyl ether was indicative of a radical fragmentation reaction. The
ratio of the fragmentation products was dependent upon a number of factors. These
included the ratio of HMPA to SmI2, the amount of methanol used, the concentration of
the reaction mixture i.e. THF volume, the number of equivalents of SmI2 added and the
reaction temperature. Dilution of the mixture favoured the C-C bond cleavage, as did an
increase in reaction temperature; at -78 °C, no C-C fragmentation product was detected
and 94% of the alcohol was recovered, while the highest percentage of C-C
fragmentation took place at 50 °C. The total product yield did, however, suffer at
elevated temperatures.
The authors suggested that the reaction is initiated by reduction of the carbon-bromine
bond to produce the oxiranylmethyl radical. Two pathways then become available. The
first is a further reduction by a second equivalent of SmI2, followed by C-O cleavage to
afford the Sm-alkoxide. The second possibility is a C-C radical fragmentation to give the
vinyl ether radical intermediate. Further reduction yields the anion which either
protonates to give the vinyl ether or it can recyclise and undergo C-O cleavage.
The timing of the anionic quenching is critical in determining the product distribution.
This is well illustrated in the graph shown below (Figure 3). If a proton source is present
in the solution it is possible to rapidly trap the carbanion formed after reduction of the CC fragmentation intermediate (radical) to yield the vinyl ether before it can recyclise and
eliminate oxygen.
rel C-C frag. %
49
45
40
35
30
25
20
15
10
5
0
0
50
100
150
200
250
300
MeOH / equiv
Figure 3. Effects of added MeOH on the relative ratio of C-C fragmentation.
The effect of added HMPA is difficult to rationalise. Intuitively one would think that the
increase in reducing potential with increasing HMPA concentration would favour the
reduction of the initially formed oxiranylmethyl radical, and yield an increasing amount
of C-O cleaved product. This is opposite to experimental facts. The relative C-C
fragmentation product increases with increasing HMPA concentration up to a maximum
at 8 equivalents of HMPA (Figure 4).
Although SmI2 with 4-5 equivalents of HMPA is considered to be most effective for the
reduction of primary alkyl radicals, the initial radical formed in this reaction is benzylic
and hence delocalised. The authors tentatively rationalised the observation by concluding
that the reduction of the carbon radical possessing bulky aryl substituents may be
sensitive to steric factors if the process involves the formation of a carbon-Sm bond
similar to that of primary alkyl radicals. As the amount of HMPA increases so does the
size of the co-ordination complex, and thus the reduction of the initial radical would slow
50
down. This would allow the C-C fragmentation pathway to compete with the further
rel C-C frag. %
reduction..
40
35
30
25
20
15
10
5
0
0
5
10
15
20
HMPA / equiv
Figure 4. Effects of added HMPA on the relative ratio of C-C fragmentation
SmI2 is able to promote the conversion of α-bromomethyl cyclic β-keto esters to the
corresponding ring-expanded one-carbon homologated γ-ketoesters in good yields. 72 The
approach involves an intramolecular samarium Barbier reaction followed by a ring
expansion sequence. The mechanism proposed is depicted below (Scheme 70).
Deuterium labelling experiments showed a large amount of deuterium incorporation in
the ring-expanded product. This lends credence to the reaction pathway that involves a
reduction of the radical intermediate, rather than hydrogen abstraction from the solvent.
With R = Me, the cyclopropanol was recovered in 98% yield. The presence of the ester
moiety is, therefore, crucial for the ring expansion that leads to a carbanion intermediate.
When the ring size of the substrate was reduced to the cyclopentane derivative, the
51
addition of HMPA and a proton source were necessary for the ring expansion. In their
absence, a naphthalene derivative was formed.
O
R
SmI2
-Br-
H
R
187
SmI2
I2SmO
R
H
189
O
188
R
SmI2
190
H+
HO
O
R
Me
R
O
186
O
O
Br
OSmI2
H+
Me
OEt
191
192
193
Scheme 70
The literature on bicyclo[n.1.0] radicals reveals a preference for stereocontrolled
exocyclic radical ring opening as opposed to the thermodynamically favoured endocyclic
ring opening. 73 Exocyclic ring opening has been achieved utilising a variety of electron
transfer techniques, including reagents such as SmI2. The selectivity of the reaction can
be altered to favour the endocyclic C-C fragmentation reaction if an appropriately
situated radical/anion stabilizing group, such as an ester moiety, is incorporated into the
substrate (Scheme 71). 74 Thus, when a solution of the cyclopropyl compound (below) is
treated with SmI2 in THF, the ring-expanded product is isolated in 44% yield. It was
found that HMPA and DMPU were ineffective as additives in increasing the yield.
However, when a proton source such as methanol was added, the reaction proceeded
52
smoothly and the yield increased two-fold. This work contrasts that of the exocyclic ring
opening, in which no stabilizing group is present on the substrate molecule.
O
O
SmI2
n CO2Bn
194
THF, MeOH
89%
n = 1, 2
n CO Bn
2
195
Scheme 71
A number of natural products and key intermediates have been elegantly synthesised
using SmI2 radical cascade methodology. (±)-Paeonilactone B has been constructed
utilising such an approach (Scheme 72). 75 The mechanism presumably involves an
intitial cyclisation of the ketyl radical onto a methylenecyclopropane unit with subsequent
‘endo’ ring opening to give the methylene cyclohexyl radical. This then cyclises onto the
pendant alkyne giving rise to a cis-fused bicyclic system. The authors found that the
addition of HMPA or DMPU was imperative for a successful reaction, with HMPA far
out performing DMPU with respect to both yield and diastereoselectivity. The observed
diasteroselectivity (cis-hydroxy vs trans-hydroxy, 10:1) is attributed to steric constraints
in the transition state. The cyclisation is thought to proceed via a chair-like transition
state, favouring the conformation in which both the bulky Sm(HMPA) enolate and the
prop-2-ynyl ether adopt pseudo-equatorial positions.
53
O
HO
HO
SmI2, t-BuOH, HMPA
O
H
+
THF, 0 °C
O
196
199
O
200
HO
I2SmO
O
I2SmO
O
O
O
O
(±)-Paeonilactone B
197
198
201
Scheme 72
A similar methodology was applied to the corresponding allyl ethers in the hope of
synthesizing paeonilactone A. When the keto-diene substrate was subjected to SmI2 /
HMPA the fragmentation reaction proceeded as before, but the yield as well as the
stereoselectivity at the newly formed chiral centre were reduced with respect to the
analogous propargyl ether (Scheme 73). Subjecting the diastereomer to the same system,
however, yielded a single diasteomeric bicyclic product (17% yield) accompanied by its
dimer as a single diastereomer in 25% yield (Scheme 74). Although this particular
reaction sequence was highly diastereoselective, it did not give the correct
stereochemistry for the desired natural product. As noted before, the use of DMPU as a
substitute for HMPA led to a loss of diastereoselectivity in the cyclisation step. Although
the cyclisations of the allyl ethers failed to provide the correct stereochemistry for
paeonilactone A, the conversion of the bicyclic ethers formed to diastereomers of the
natural product were investigated. This allowed for the efficient and stereoselective
synthesis of (±)-6-epi-paeonilactone A.
54
O
SmI2, t-BuOH, HMPA
O
H
HO
HO
THF, 0 °C
H
H
+
H
H
O
1:1 (35%)
O
203
202
204
Scheme 73
O
SmI2, ButOH, HMPA
O
H
205
THF, 0 °C
OH
HO
HO
H
H
H
H +
H
O
206
17%
O
O
H
207
25%
Scheme 74
Walborski and Topolski 76 studied the reaction of chiral cyclopropyl halides with SmI2 in
the presence of HMPA. The major product was the racemic reduced cyclopropyl
compound, with the alkene and dimeric product being only minor components (Scheme
75). By carrying out these reactions in the presence of deuterated methanol, the authors
were able to deduce which products were formed via a radical process and which were
formed via an anionic sequence. Their experiments showed only a 15 % deuterium
incorporation into the cyclopropyl derivative, implying that ring opening or H-atom
abstraction from the solvent occurs before the radical can be further reduced. The alkene,
however, possessed one deuterium atom per molecule, proving that it was formed after
the radical had been reduced by a second equivalent of SmI2.
55
Ph
Ph
CH3
MeOD
CH3
Ph
[SmI2]+
210
Ph
211
S-H
SmI2
Ph
Ph
CH3
Br
Ph
CH3
Ph
(S)-(+)
209
208
Ph
CH3
SmI2
Ph
CH2D
MeOD
CH3
Ph
Ph
213
CH3
Ph
CH3
Ph
CH2 CH2
212
Ph
Ph
214
Scheme 75
SmI2 can also induce the regioselective cleavage of phenylsulfonyl activated cyclopropyl
ketones. 77 The cleavage of these cyclopropanes followed by β-elimination of
phenylsulfonyl radical leading to β,γ-unsaturated ketones has been demonstrated (Scheme
76).
Ph H H Ph
PhO2S
H
Ph
H
O
H
Ph H H Ph
SO2Ph + PhO2S
H
H
SmI2, HMPA,
tBuOH
215
PhO2S
H
O
H
216
O
Ph + Ph
SO2Ph
H
O
Ph + Ph
Ph
O
217
67%
218
9%
Scheme 76
219
3%
56
The SmI2-Fe(DBM)3 [tris(dibenzoylmethido)iron(III)] reagent system has been used to
successfully promote the ring opening reaction of a number of cyclopropane-1,1dicarboxylic esters. 78 Excellent yields were achieved in a relatively short period of time
with a number of different ester moieties. When the same reactions were carried out at
reflux temperature in the presence of an aliphatic ketone, the respective 5-pentanolide
derivatives were isolated (Scheme 77). The addition of aldehydes or aromatic ketones
resulted in a significant amount of pinacol products with low yields of the desired 5pentanolides. S,S'-Diphenyl cyclopropane-1,1-dicarbothioate was allowed to react with
carbonyl compounds under similar conditions, giving δ-hydroxy esters in moderate yield.
R1
CO2Me
R
H
+
R1COR2
SmI2, Fe(DBM)3 (4 mol%)
CO2Me
O
CO2Me
R
THF,
220
O
R2
222
221
Scheme 77
Yamashita and co-workers have expanded on the ring opening of cyclopropanecarboxylic
esters and cyclopropane-1,1-dicarboxylic esters with a SmI2-HMPA-THF system in the
presence of t-BuOH as a proton source to give 4-substituted butyric esters and (2substituted ethyl)malonic esters (Scheme 78). 79
H
CO2Et
H
CO2Et
R
SmI2
THF,
223
R
R
H
COOEt
COOEt
COOEt
COOEt
224
Scheme 78
57
In the absence of a proton source, however, they were able to reductively dimerise
several 2-substituted cyclopropane-1,1-dicarboxylic esters. 80 The initial yields were poor,
but were increased by the omission of HMPA from the system. The yield was further
improved by conducting the reactions in a solution of SmI2-THF under reflux.
1.5.3 1,4-Diketones in non-strained and strained systems
Hoffmann and co-workers 81 have recently described a 1,4-pinacolisation methodology
utilising SmI2 that is useful for the production of highly strained systems containing a
1,2-cyclobutanediol moiety. Ghosh anticipated that this methodology would be useful for
gaining easy access to [3.3.2]propellanes. Thus, bicyclo[2.2.1]heptane derivatives bearing
a 1,4-dicarbonyl moiety were prepared and subjected to reaction with SmI2. 82 Contrary to
expectations of cyclisation of the substrate, fragmentation to macrocyclic compounds was
observed (Scheme 79). Although the pinacol reaction is a possible reaction outcome, it
would lead to an increase in strain of an already strained norbornene system: the more
facile C-C fragmentation pathway is, therefore, followed. Relieving some ring strain by
prior reduction of the double bond did not alter the outcome of the reaction. The addition
of HMPA might have been a deciding factor in this reaction.
MeOC
H
SmI2, THF
COMe HMPA, t-BuOH
COMe
100%
225
H
MeOC
226
Scheme 79
58
Camps has recently reported a similar experience of the fragmentation of a 1,4-diketone
while trying the Hoffmann pinacol methodology. 83 Upon subjecting the strained
bisnoradamantane system to SmI2, it readily underwent fragmentation to give a mixture
of three stereoisomeric bicyclic diketones in 80% yield (Scheme 80). A mixture of
stereoisomeric alcohols (16% yield) derived from further reduction of the diketones was
also isolated.
Me
Me
t-Bu-OC
Me
CO-t-Bu
t-Bu
HO
SmI2,
THF
227
Me
t-Bu
OH
228
Me
t-Bu-OC
CO-t-Bu + diastereomers
Me
230
229
Me
+
t-Bu-OC
Me
H
C OH
t-Bu
231
Scheme 80
A molecular mechanics (MM2 and MM3) investigation was carried out on the substrate,
the possible cyclobutanediol isomers, and the fragmented diketones giving the formation
enthalpies as well as the strain energies of the individual molecules. The calculations
confirmed natural intuition, showing an enormous increase in strain energy for the
pinacol products, while the strain energy of the products after fragmentation was
considerably reduced. The authors went further and stated that it was reasonable to
59
assume that the transition-state for the conversion of the diketyl radical derived from the
bisnoradamantane diketone to the bis-enolate derived from the fragmented product is of
much lower energy than the corresponding transition-state for its conversion to the
diolate derived from the pinacol product. The fragmentation pathway via the diketyl
radical is therefore more facile than that of its pinacol counterpart.
1.6 Three and Four Membered Rings
One of the ways in which SmI2 is made is the reaction of samarium metal with
diiodomethane. The intermediate formed is ‘ISmCH2I’ which undergoes α-elimination
generating SmI2 and methylene. This enables one to trap the carbenoid intermediate with
olefins thus producing cyclopropanes and providing a useful alternative to the traditional
Simmons-Smith procedure. In practice, a more economical reagent to use for the
cyclopropanation is the Sm(Hg) amalgam. The SmI2-mediated reaction offers enhanced
chemoselectivity and often higher diastereoselectivity than in classical Simmons-Smith
reactions (Scheme 81). 84
CF3
OH
Sm(Hg), CH2I2
OH
C6H13
THF
232
CF3
H
C6H13
H
233
Scheme 81
Enolates are inert to cyclopropanation with the Simmons-Smith reagent system and form
α-methylated ketones instead. By contrast generation of the kinetic enolate with LDA
60
followed by the addition of SmI2/CH2I2 generates cyclopropanes in good yield (Scheme
82). 85
O
OH
1. LDA, THF
2. SmI2, CH2I2
234
235
Scheme 82
The methodology can be extended to α-halo ketones. The enolate is generated in situ by
reaction with SmI2 and cyclopropanation follows (Scheme 83). 86
O
HO
3 CH2I2, 2 SmI2
Br
THF
236
237
Scheme 83
It is possible to directly convert carboxylic acid esters to cyclopropanols by reacting them
with diiodomethane and SmI2 (generated in situ). The reaction proceeds via a
nucleophilic acyl substitution reaction forming the corresponding iodomethyl ketones.
Reduction of this α-hetrosubstituted ketone by SmI2 generates the enolate, which is
finally trapped with the carbenoid intermediate to give the cyclopropane product (Scheme
84).85
O
R
O
CH2I2
OR'
238
Sm
OSmI2
SmI2
R
CH2I
239
R
CH2
240
Scheme 84
CH2I2
OH
R
Sm
241
61
The samarium Barbier reaction of nonracemic β-chloro-substituted amides provides high
yields of the corresponding cyclopropanols with good diastereoselectivity. 87 The Sm(III)
by-product acts an effective Lewis acid and controls the stereoselectivity of the reaction
by chelating the functional groups (Scheme 85).
O
O
O
O
2 SmI2
Cl
N
THF, HMPA
O
HO
N
CH2Ph
CH2Ph
242
243
Scheme 85
One of the more popular applications of the Barbier reaction is the synthesis of bicyclic
alcohols (Scheme 86). The construction of cyclopropanols is accomplished by treating αtosyloxymethyl cyclohexanones with SmI2. 88
O
OTs
HO
2 SmI2
THF
244
245
Scheme 86
A similar reaction was attempted to produce a cyclobutanol derivative but the cyclisation
did not occur. Rather, a reductive β-elimination took place generating a samarium enolate
(Scheme 87).87
62
I
O
O
OSmI2
SmI2
THF
246
248
247
Scheme 87
The formation of cyclopropanols in exceedingly good yields from β-bromo ketones or
aldehydes is particularly impressive. 89 The substrate is usually prepared in situ from 3bromopropionates via a Grignard reaction. The analogous reaction to produce
cyclobutanes has also been attempted but much poorer yields were obtained (Scheme 88).
Br
OEt
1. RMgX
n
O
249
n
2. SmI2, THF-HMPA
R
OH
250
n = 1 Yield = 70-99%
=2
= 5%
Scheme 88
Nucleophilic acyl substitution reactions with esters are, however, amenable to the
synthesis of four membered rings. The reactions are accomplished under mild conditions
upon treatment with SmI2 (Scheme 89). 90
I
CO2Et
Ph
O
2 SmI2
THF, cat. Fe(III)
74%
Ph
252
251
Scheme 89
63
Cyclopropanols are readily accessible via a sequential process involving an acyl radical
cyclisation followed by an intermolecular Barbier reaction. 91 Thus 2-allyloxybenzoyl
chloride reacts with SmI2 generating an acyl radical. This undergoes intramolecular
addition to the double bond, thereby forming a new radical. The radical is further reduced
by a second equivalent of SmI2 to form the organosamarium intermediate, which reacts
with the ketone giving the desired cyclopropanol (Scheme 90).
O
HO
2 SmI2
Cl
THF
O
253
O
254
Scheme 90
In a similar vein, sequential processes have been developed to allow access to bicyclic
structures containing the cyclobutyl moiety. 92 These protocols are also commonly used in
the synthesis of heterocyclic systems (Scheme 91). Both activated alkynes and
unactivated olefins can be used as radical traps in the reaction.
O
I
EtO
O
SmI2
SiMe3
THF-HMPA
255
OH
O
256
Scheme 91
SiMe3
64
Another intramolecular Barbier reaction allows access to cyclobutanols with an alkenyl
moiety. The same protocol was used as in the previous example except that the substrate
is set up for β-elimination after the bicyclisation has occurred (Scheme 92). 93
HO
O
EtO
O
Br
SmI2
O
O
HO
THF-HMPA
259
258
257
Scheme 92
Another way one can envisage making small ring systems is via radical cyclisation of a
suitable intermediate onto the C-C double bond of α,β-unsaturated esters.
Guibé and co-workers have shown that δ-iodo and δ-bromo-α,β-unsaturated esters with
various substituents at the β- and γ-positions readily cyclise to form cyclopropane
compounds under the reducing power of SmI2 and in the presence of a proton source
(Scheme 93). 94
CO2Bn
I
SmI2, THF
CO2Bn
t-BuOH
261
260
Scheme 93
The attempted SmI2-mediated cyclisation reaction of a δ-iodo-α,β-unsaturated ester to
afford a cyclobutanol failed, giving the reduced product in its stead (Scheme 94). 95
65
O
O
OEt
1.5 SmI2, cat. NiI2
OEt
THF, t-BuOH
I
263
262
Scheme 94
In order to obtain the desired cyclobutanol product in these types of reactions it is
essential to make use of the gem-disubstitution effect. 96 This was well illustrated by
Weinges and co-workers who successfully cyclised a gem-dimethyl-γ,δ-unsaturated
aldehyde to the corresponding cyclobutanol in the presence of SmI2 (Scheme 95). 97 This
work was ellaborated on by Procter and co-workers who managed to improve the reaction
conditions and yields but were unable to obtain any cyclobutanol product without some
sort of disubstitution in the substrate. 98
BnO
CHO
SmI2
CO2Et
BnO
OH
THF-HMPA
264
CO2Et
265
Scheme 95
Yet another method of synthesising cyclobutane derivatives is that of the pinacol
coupling reaction of 1,4-diketone substrates mediated by SmI2. This methodology,
developed by Hoffmann and co-workers, allows for easy access into a variety of
substituted 1,2-cyclobutanediol derivatives. 99 The beauty of the reaction stems from the
66
fact that no gem-disubstitution is necessary and the reaction requires no co-solvents or
additives (Scheme 96).
O
R2
R1
O
HO
SmI2
OH
R1
THF
266
R2
267
Scheme 96
The pinacol reaction is also very powerful and enables the formation of extremely ring
strained compounds (Scheme 97). 100
O R
R
SmI2
H
THF
OH OH
O
269
268
Scheme 97
1.7 Conclusions
1.7.1 Fragmentations
SmI2, either in the presence or absence of co-solvents, promotes manifold reactions,
many of which are fragmentation reactions. In a number of cases, the fragmentation
methodology has been put to good use in the preparation of highly functionalised
products, which often are natural products or those that possess physiological activity.
The fragmentation protocol has been found useful in a variety of transformations, and has
67
been effectively applied to sequential reactions in which one or more transformations
follow the initial fragmentation step.
SmI2 has been found to be the reductant of choice for many reductive fragmentation
reactions, and generally affords yields higher than those provided by other reducing
agents, and in most instances affords higher chemical and stereochemical yields.
This being the case, SmI2 should continue to find application in many synthetic
sequences in more and more laboratories engaged in organic synthesis.
1.7.2 Small Ring Systems
Samarium diiodide has been used to synthesise a number of small ring systems with good
stereoselectivity. The use of carbohydrates as precursors to these compounds and hence
the production of valuable, chiral small ring systems is, however, lacking. The
methodologies presented in the literature have also not been tested on complex molecules
such as sugars where the potential for fragmentation reactions is elevated. The current
synthetic routes to three and four membered ring systems lack the capacity to produce
highly substituted end products. These areas require attention and will be addressed in
this manuscript.
1.8 References
1
(a) Girard, P., Namy, J.L., Kagan, H.B. J. Am. Chem. Soc. 1980, 102, 2693. (b)
Molander, G.A. Chem. Rev. 1992, 92, 29. (c) Sasaki, M., Collin, J., Kagan, H.B. New J.
Chem. 1992, 16, 89.
68
2
(a) Inanaga, J., Ishikawa, M., Yamaguchi, M. Chem. Lett. 1987, 1485. (b) Ruder, S.M.
Tetrahedron Lett. 1992, 33, 2621. (c) Cabri, W., Candiani, I., Colombo, M., Franzoi, L.,
Bedeschi, A. Tetrahedron Lett. 1995, 36, 949. (d) Hasegawa, E., Curran, D.P. J. Org.
Chem. 1993, 58, 5008.
3
(a) Ogawa, A., Takami, N., Sekiguchi, M., Ryu, I., Kambe, N., Sonoda, N. J. Am.
Chem. Soc. 1992, 114, 8729. (b) Ogawa, A., Nanke, T., Takami, N., Sumino, Y., Ryu, I.,
Sonoda, N. Chem. Lett. 1994, 379.
4
Lu, X., Ma, S., Zhu, J. Tetrahedron Lett. 1988, 29, 5129.
5
Bennet, S.M., Larouche, D. Synlett 1991, 805.
6
(a) Fevig, T.L., Elliot, R.L., Curran, D.P. J. Am. Chem. Soc. 1988, 110, 5064. (b)
Molander, G.A., Kenny, C. J. Am. Chem. Soc. 1989, 111, 8236. (c) Molander, G.A.,
Harring, L.S. J. Org. Chem. 1990, 55, 6171. (d) Inanaga, J., Ujikawa, O., Yamaguchi, M.
Tetrahedron Lett. 1991, 32, 1737.
7
Fukuzawa, S.,Tsuchimoto, T. Synlett 1993, 803.
8
Aurrecoechea, J.M., Fernandez-Acebes, A. Tetrahedron Lett. 1993, 34, 549.
9
Batey, R.A., Motherwell, W.B. Tetrahedron Lett. 1991, 32, 6649.
10
(a) Sturino, C.F., Fallis, A.G. J. Am. Chem. Soc. 1994, 116, 7447. (b) Sturino, C.F.,
Fallis, A.G. J. Org. Chem. 1994, 59, 6514
11
(a) Fukuzawa, S, Nakanishi, A., Fujinami, T., Sakai, S. J. Chem. Soc. Chem. Commun.
1986, 624. (b) Fukuzawa, S, Nakanishi, A., Fujinami, T., Sakai, S. J. Chem. Soc. Perkin
Trans. 1 1988, 1669. (c) Otsubo, K., Inanaga, J., Yamaguchi, M. Tetrahedron Lett. 1986,
27, 5763.
12
Ujikawa, O., Inanaga, J., Yamaguchi, M. Tetrahedron Lett. 1989, 30, 2837.
69
13
Inanaga, J., Katsuki, J., Ujikawa, O., Yamaguchi, M. Tetrahedron Lett. 1991, 32, 4921.
14
Kito, M., Sakai, T., Yamada, K., Matsuda, F., Shirahama, H. Synlett 1993, 158.
15
Baldwin, J.E., Turner, S.C.M., Moloney, M.G. Tetrahedron 1994, 50, 9411.
16
Kan, T., Nara, S., Ito, S., Matsuda, F., Shirahama, H. J. Org. Chem. 1994, 59, 5111.
17
Shim, S.C., Hwang, J.-T., Kang, H.-Y., Chang, M.H. Tetrahedron Lett. 1990, 31,
4765.
18
Molander, G.A., McKie, J.A. J. Org. Chem. 1995, 60, 872.
19
Molander, G.A., McKie, J.A. J. Org. Chem. 1992, 57, 3132.
20
Bannai, K., Tomimori, K., Kato, Y., Kurozumi, S., Nyori, R. Tetrahedron 1990, 46,
6689.
21
Gillmann, T. Tetrahedron Lett. 1993, 34, 607.
22
(a) Shiue, J.-S., Lin, C.-C., Fang, J.-M. Tetrahedron Lett. 1993, 34, 335. (b) Kise, N.,
Suzumoto, T., Shono, J. J. Org. Chem. 1994, 59, 1407.
23
(a) Namy, J.L., Souppe, J., Kagan, H.B. Tetrahedron Lett. 1983, 24, 765. (b) Fürstner,
A., Csuk, R., Rohrer, C., Weidmann, H.J. J. Chem. Soc., Perkin Trans. 1 1988, 1729. (c)
Akane, N., Kanagawa, Y., Nishiyama, Y., Ishii, Y. Chem. Lett. 1992, 2431.
24
(a) Enholm, E.J., Forbes, D.C., Holub, D.P. Synthetic Commun. 1990, 20, 981. (b)
Imamoto, T., Nishimura, S. Chem. Lett. 1990, 1141.
25
(a) Chiara, J.L., Cabri, W., Hanessian, S. Tetrahedron Lett. 1991, 32, 1125. (b) Chiara,
J.L., Martin-Lomas, M. Tetrahedron Lett. 1994, 35, 2969. (c) Uenishi, J., Masuda, S.,
Wakabayashi, S. Tetrahedron Lett. 1991, 32, 5097.
26
Molander, G.A., Kenny, C.J. J. Org. Chem. 1988, 53, 2132.
27
Clayden, J., Julia, M. J. Chem. Soc., Chem. Commun. 1994, 2261.
70
28
Molander, G.A., Etter, J.B. Synth. Commun. 1987, 17, 901.
29
Molander, G.A., Etter, J.B., Zinke, P.W. J. Am. Chem. Soc., 1987, 109, 453.
30
Suginome, H., Yamada, S. Tetrahedron Lett. 1987, 28, 3963.
31
(a) Sosnowski, J.J., Danaher, E.B., Murray, R.K. Jr. J. Org. Chem. 1985, 50, 2759. (b)
Molander, G.A., McKie, J.A. J. Org. Chem. 1991, 56, 4112.
32
Kagan, H.B., Namy, J.L., Girard, P. Tetrahedron 1981, 37, 175, Suppl. 1.
33
Zhang, Y., Liu, T., Lin, R. Synth. Commun. 1988, 18, 2003.
34
Arime, T., Kato, N., Komadate, F., Seegusa, H., Mori, N. Synthetic Commun. 1994, 24,
3315.
35
Molander, G.A., Etter, J.B., Harring, L.S., Thorel, P.-J. J. Am. Chem. Soc. 1991, 113,
8036.
36
Molander, G.A., McKie, J.A. J. Org. Chem. 1993, 58, 7216.
37
Fevig, T.L., Elliot, R.L., Curran, D.P. J. Am. Chem. Soc. 1988, 110, 5064.
38
Curran, D.P., Totleben, M.J. J. Am. Chem. Soc. 1992, 114, 6050.
39
Molander, G.A., Hahn, G. J. Org. Chem. 1986, 51, 1135.
40
Lannoye, G.,Cook, J.M. Tetrahedron Lett. 1988, 29, 171.
41
Molander, G.A. Chem. Rev. 1992, 92, 29.
42
(a) Hasegawa, E., Curran, D.P. J. Org. Chem. 1993, 58, 5008. (b) Curran, D.P., Wolin,
R.L. Synlett. 1991, 317.
43
Hojo, M., Aihara, H., Hosomi, A. J. Am. Chem. Soc. 1996, 118, 3533.
44
Shabangi, M., Flowers, R.A., II Tetrahedron Lett. 1997, 38, 1137.
45
Hou, Z., Wakatsuki, Y. J. Chem. Soc. Chem. Commun. 1994, 1205.
46
Hasegawa, E., Curran, D.P. Tetrahedron Lett. 1993, 34, 1717.
71
47
(a) Fevig, T.L., Elliot, R.L., Curran, D.P. J. Am. Chem. Soc. 1988, 110, 5064. (b)
Molander, G.A., Harring, L.S. J. Org. Chem. 1990, 55, 6171. (c) Shiue, J.-S., Lin, C.-C.,
Fang, J.-M. Tetrahedron Lett. 1993, 34, 335.
48
Hou, Z., Zhang, Y., Wakatsuki, Y. Bull. Chem. Soc. Jpn. 1997, 70, 149.
49
Evans, W.J., Gummersheimer, T.S., Ziller, J.W. J. Am. Chem. Soc. 1995, 117, 8999.
50
Shotwell, J.B., Sealy, J.M., Flowers, R.A. II J. Org. Chem. 1999, 64, 5251.
51
Ananthanarayan, T.P., Gallagher, T. Magnus, P. J. Chem. Soc., Chem. Commun. 1982,
709.
52
Honda, T., Naito, K., Yamane, S.-i., Suzuki, Y. J. Chem. Soc., Chem. Commun. 1992,
1218.
53
Molander, G.A., McKie, J.A. J. Org. Chem. 1991, 56, 4112.
54
(a) Honda, T., Yamane, S.-i., Naito, K., Suzuki, Y. Heterocycles 1994, 37, 515. (b)
Honda, T., Ishikawa, F., Yamane, S.-i. J. Chem. Soc., Chem. Commun. 1994, 499. (c)
Honda, T., Yamane, S.-I., Naito, K., Suzuki, Y. Heterocycles 1995, 40, 301. (d) Honda,
T., Ishikawa, F., Yamane, S.-I. J. Chem. Soc., Perkin Trans. 1 1996, 1125. (e) Honda, T.,
Yamane, S.-I., Ishikawa, F., Katoh, M. Tetrahedron 1996, 52, 12177. (f) Honda, T.,
Katoh, M., Yamane, S.-I. J. Chem. Soc., Perkin Trans. 1 1996, 2291 .
55
56
Honda, T., Ishikawa, F. J. Org. Chem. 1999, 64, 5542.
Molander, G.A., Etter, J.B., Zinke, P.W. J. Am. Chem. Soc 1987, 109, 453.
57
Kang, H.-Y., Sang Hong, W. Tetrahedron Lett. 1995, 42, 7661.
58
Meng, Q., Hesse, M. "Topics in Current Chemistry", 1992, Vol. 161, 107, Springer-
Verlag.
59
Goulaouic-Dubois, C., Guggisberg, A., Hesse, A. Tetrahedron 1995, 51, 12035.
72
60
Lange, G.L., Gottardo, C. Tetrahedron Lett. 1994, 35, 6607.
61
Lange, G.L., Gottardo, C., Merica, A. J. Org. Chem. 1999, 64, 6738.
62
Lange, G.L., Merica, A. Tetrahedron Lett. 1999, 40, 7897.
63
Lange, G.L., Furlan, L., MacKinnon, M.C. Tetrahedron Lett. 1998, 39, 5489.
64
Comins, D.L., Zheng, X. J. Chem. Soc., Chem. Commun. 1994, 2681.
65
(a) Batey, R.A., Motherwell, W.B. Tetrahedron Lett. 1991, 32, 6211. (b) Batey, R.A.,
Harling, J.D., Motherwell, W.B. Tetrahedron 1996, 52, 11421.
66
Molander, G.A., Alonso-Alija, C. Tetrahedron 1997, 53, 8067.
67
Hwang, J.-T., Liao, C.-C. Tetrahedron Lett. 1991, 32, 6583.
68
(a) Nivlet, A., Le Guen, V., Dechoux, L., Le Gall, T., Mioskowski, C. Tetrahedron
Lett. 1998, 39, 2115. (b) Nivlet, A., Dechoux, L., Le Gall, T., Mioskowski, C. Eur. J.
Org. Chem. 1999, 3251.
69
Beerli, R., Brunner, E.J., Borschberg, H.-J. Tetrahedron Lett. 1992, 33, 6449.
70
Grob, A. Angew. Chem. Int. Ed. Engl. 1965, 8, 535.
71
Hasegawa, E., Takahashi, M., Horaguchi, T. Tetrahedron Lett. 1995, 36, 5215.
72
Hasegawa, E., Kitazume, T., Suzuki, K., Tosaka, E. Tetrahedron Lett. 1998, 39, 4059.
73
(a) Cristol, S., Barbour, R.V. J. Am. Chem. Soc. 1968, 90, 2832. (b) Beckwith, A.L.J.,
Philliou, G.J. J. Chem. Soc. Chem Commun. 1971, 658. (c) Beckwith, A.L.J., Philliou,
G.J. Aust. J. Chem. 1976, 29, 123. (d) Friedrich, E.C., Holmstead, R.L. J. Org. Chem.
1971, 36, 971. (e) Ingold, K.U., Walton, J.C., Acc. Chem. Res. 1986, 19, 72.
74
Lee, P.H., Lee, J. Tetrahedron Lett. 1998, 39, 7889.
75
a) Boffey, R.J., Whittingham, W.G., Kilburn, J.D. J. Chem. Soc., Perkin Trans. 1 2001,
487. (b) Boffey, R.J., Santagostino, M., Whittingham, W.G., Kilburn, J.D. J. Chem. Soc.,
73
Chem. Commun. 1998, 1875. (c)
Bofey, R.J., Whittingham, W.G., Kilburn, J.D.
Tetrahedron Lett. 1999, 40, 5625.
76
Walborsky, H.M., Topolski, M., J. Org. Chem. 1992, 57, 370.
77
Reutrakul, V., Saeeng, R., Pohmakotr, M., Kongsaeree, P. Tetrahedron Lett. 1999, 40,
1019.
78
Imamoto, T., Hatajima, T., Yoshizawa, T. Tetrahedron Lett. 1994, 35, 7805.
79
(a) Yamashita, M., Okuyama, K., Ohhara, T., Kawasaki, I., Ohto, S. Chem. Pharm.
Bull. 1995, 43, 708. (b) Yamashita, M., Okuyama, K., Ohhara, T., Kawasaki,I., Sakai, K.,
Nakata, N., Kawabe, T., Kusumoto, M., Ohto, S. Chem. Pharm. Bull. 1995, 43, 2075.
80
Yamashita, M., Okuyama, K., Ohhara, T., Kawasaki, T., Ohta, S. Synlett. 1996, 547.
81
(a) Hoffmann, H.M.R., Munnich, I., Nowitzki, O., Stucke, H., Williams, D.J.
Tetrahedron 1996, 52, 11783. (b)Nowitzki, O., Munnich, I., Stucke, H., Hoffmann,
H.M.R. Tetrahedron 1996, 52, 11799.
82
Haque, A., Ghosh, S. J. Chem. Soc., Chem. Commun. 1997, 2039.
83
Camps, P., Lukach, A.E., Vázquez, S. Tetrahedron 2001, 57, 2419.
84
(a) Molander, G.A., Etter, J.B. J. Org. Chem. 1987, 52, 3942. (b) Molander, G.A.,
Harring, L.S. J. Org. Chem. 1989, 54, 3525. (c) Yamazaki, T., Lin, J.T., Takeda, M.,
Kitazume, T. Tetrahedron: Assymetry 1990, 1, 351.
85
Immamoto, T., Takiyama, N. Tetrahedron Lett. 1987, 28, 1307.
86
Immamoto, T., Hatajima, T., Takiyama, N., Takeyama, T., Kamiya, Y., Yoshizawa, T.
J. Chem. Soc., Perkin Trans. 1 1991, 3127.
87
Fadel, A. Tetrahedron: Assymetry 1994, 5, 531.
88
Berks, A. Ph.D. Thesis, University of Colorado, Boulder, 1988.
74
89
Fukuzawa, S.-i., Furuya`, H., Tsuchimoto, T. Tetrahedron 1996, 52, 1953.
90
Molander, G.A., McKie, J.A. J. Org. Chem. 1993, 58, 7216.
91
Sasaki, M., Collin, J., Kagan, H.B. Tetrahedron Lett. 1988, 29, 6105.
92
Molander, G.A., Harris, C.R. Tetrahedron Lett. 1998, 54, 3321.
93
Molander, G.A., Harris, C.R. J. Org. Chem. 1998, 63, 4374.
94
David, H., Afonso, C., Bonin, M., Doisneau, G., Guillerez, M.-G., Guibé, F.
Tetrahedron Lett. 1999, 8557.
95
Molander, G.A., Harris, C.R. J. Org. Chem. 1997, 62, 7418.
96
(a) Jung, M.E., Marquez, R., Houk, K.N. Tetrahedron lett. 1999, 40, 2661. (b) Park, S.-
U., Varick, T.R., Newcomb, M. Tetrahedron Lett. 1990, 31, 2975.
97
Weinges, K., Schmidbauer, S.B., Schick, H. Chem. Ber. 1994, 127, 1305.
98
(a) Johnston, D., McCusker, C.M., Procter, D.J. Tetrahedron Lett. 1999, 4913. (b) )
Johnston, D., McCusker, C.F., Muir, K., Procter, D.J. J. Chem. Soc., Perkin Trans. 1
2000, 681.
99
Hoffmann, H.M.R., Munnich, I., Nowitzki, O., Stucke, H., Williams, D.J. Tetrahedron
1996, 52, 11783.
100
Nowitzki, O., Munnich, I., Stucke, H., Hoffmann, H.M.R. Tetrahedron 1996, 52,
11799.