Xinfang Xu1, Michael P Doyle. 1. Department of Chemistry and Biochemistry, University of Maryland , College Park, Maryland 20742, United States.
Abstract
The combination of two or more unsaturated structural units to form cyclic organic compounds is commonly referred to as cycloaddition, and the combination of two unsaturated structural units that forms a six-membered ring is formally either a [5 + 1]-, [4 + 2]-, [2 + 2 + 2]-, or [3 + 3]-cycloaddition. Occurring as concerted or stepwise processes, cycloaddition reactions are among the most useful synthetic constructions in organic chemistry. Of these transformations, the concerted [4 + 2]-cycloaddition, the Diels-Alder reaction, is by far the best known and most widely applied. However, although symmetry disallowed as a concerted process and lacking certifiable examples until recently, stepwise [3 + 3]-cycloadditions offer advantages for the synthesis of a substantial variety of heterocyclic compounds, and they are receiving considerable attention. In this Account, we present the development of stepwise [3 + 3]-cycloaddition reactions from virtual invisibility in the 1990s to a rapidly growing synthetic methodology today, involving organocatalysis or transition metal catalysis. With origins in organometallic or vinyliminium ion chemistry, this area has blossomed into a viable synthetic transformation for the construction of six-membered heterocyclic compounds containing one or more heteroatoms. The development of [3 + 3]-cycloaddition transformations has been achieved through identification of suitable and compatible reactive dipolar adducts and stable dipoles. The reactive dipolar species is an energetic dipolar intermediate that is optimally formed catalytically in the reaction. The stepwise process occurs with the reactive dipolar adduct reacting as an electrophile or as a nucleophile to form the first covalent bond, and this association provides entropic assistance for the construction of the second covalent bond and the overall formal [3 + 3]-cycloaddition. Organocatalysis is well developed for both inter- and intramolecular synthetic transformations, but the potential of transition metal catalysis for [3 + 3]-cycloaddition has only recently emerged. The key to the rapid development of the transition metal-based methodology has been recognition that certain catalytically generated vinylcarbenes are effective dipolar adducts for reactions with stable dipolar compounds, including aryl and vinyl ylides. In particular, metallo-enolcarbenes that are generated catalytically from conveniently prepared stable enoldiazoacetates or from donor-acceptor cyclopropenes are highly effective dipolar adducts for [3 + 3]-cycloaddition. The electron-donating oxygen of the silyl ether enhances electrophilic ring closure to the metal-bound carbon of the initial adduct from vinylogous addition, and this enhancement inhibits the alternative [3 + 2]-cycloaddition across the carbon-carbon double bond of the vinylcarbene. Catalytically generated metallo-enolcarbenes react under mild conditions with a broad spectrum of compatible stable dipoles, including nitrones, azomethine imines, ylides, and certain covalent precursors of stable dipoles, to form [3 + 3]-cycloaddition products having the β-ketoester functionality (in dihydrooxazines, tetrahydropyridazines, pyrazolidinone and pyraxole derivatives, dihydroquinolines, and quinolizidines, for example) in high yield. Two ways to access these metallo-enolcarbenes, either by dinitrogen extrusion from enoldiazoacetate esters or by rearrangement of donor-acceptor cyclopropenes, enhance the versatility of the process. The [3 + 3]-cycloaddition methodology is a complementary strategy to [4 + 2]-cycloaddition for the synthesis of heterocyclic compounds having six-membered rings. High levels of enantioselectivity are obtained with the use of chiral ligands on transition metal catalysts that include those on dirhodium(II) and silver(I).
The combination of two or more unsaturated structural units to form cyclic organic compounds is commonly referred to as cycloaddition, and the combination of two unsaturated structural units that forms a six-membered ring is formally either a [5 + 1]-, [4 + 2]-, [2 + 2 + 2]-, or [3 + 3]-cycloaddition. Occurring as concerted or stepwise processes, cycloaddition reactions are among the most useful synthetic constructions in organic chemistry. Of these transformations, the concerted [4 + 2]-cycloaddition, the Diels-Alder reaction, is by far the best known and most widely applied. However, although symmetry disallowed as a concerted process and lacking certifiable examples until recently, stepwise [3 + 3]-cycloadditions offer advantages for the synthesis of a substantial variety of heterocyclic compounds, and they are receiving considerable attention. In this Account, we present the development of stepwise [3 + 3]-cycloaddition reactions from virtual invisibility in the 1990s to a rapidly growing synthetic methodology today, involving organocatalysis or transition metal catalysis. With origins in organometallic or vinyliminium ion chemistry, this area has blossomed into a viable synthetic transformation for the construction of six-membered heterocyclic compounds containing one or more heteroatoms. The development of [3 + 3]-cycloaddition transformations has been achieved through identification of suitable and compatible reactive dipolar adducts and stable dipoles. The reactive dipolar species is an energetic dipolar intermediate that is optimally formed catalytically in the reaction. The stepwise process occurs with the reactive dipolar adduct reacting as an electrophile or as a nucleophile to form the first covalent bond, and this association provides entropic assistance for the construction of the second covalent bond and the overall formal [3 + 3]-cycloaddition. Organocatalysis is well developed for both inter- and intramolecular synthetic transformations, but the potential of transition metal catalysis for [3 + 3]-cycloaddition has only recently emerged. The key to the rapid development of the transition metal-based methodology has been recognition that certain catalytically generated vinylcarbenes are effective dipolar adducts for reactions with stable dipolar compounds, including aryl and vinyl ylides. In particular, metallo-enolcarbenes that are generated catalytically from conveniently prepared stable enoldiazoacetates or from donor-acceptor cyclopropenes are highly effective dipolar adducts for [3 + 3]-cycloaddition. The electron-donating oxygen of the silyl ether enhances electrophilic ring closure to the metal-bound carbon of the initial adduct from vinylogous addition, and this enhancement inhibits the alternative [3 + 2]-cycloaddition across the carbon-carbon double bond of the vinylcarbene. Catalytically generated metallo-enolcarbenes react under mild conditions with a broad spectrum of compatible stable dipoles, including nitrones, azomethine imines, ylides, and certain covalent precursors of stable dipoles, to form [3 + 3]-cycloaddition products having the β-ketoester functionality (in dihydrooxazines, tetrahydropyridazines, pyrazolidinone and pyraxole derivatives, dihydroquinolines, and quinolizidines, for example) in high yield. Two ways to access these metallo-enolcarbenes, either by dinitrogen extrusion from enoldiazoacetate esters or by rearrangement of donor-acceptor cyclopropenes, enhance the versatility of the process. The [3 + 3]-cycloaddition methodology is a complementary strategy to [4 + 2]-cycloaddition for the synthesis of heterocyclic compounds having six-membered rings. High levels of enantioselectivity are obtained with the use of chiral ligands on transition metal catalysts that include those on dirhodium(II) and silver(I).
Two reviews in 1996 and 1997[1,2] reported that [3 + 3]-cycloadditions
were rare and that very few reports of these processes existed. They
included palladium-catalyzed generation of trimethylenemethanes as
dipolar adducts, which underwent stepwise cycloaddition with stable
dipoles that included β-diketones and β-ketoesters,[3] as well as aziridines.[4] Lithiated alkoxyallenes were subsequently established in reactions
with aldonitrones to provide an efficient [3 + 3]-methodology to oxazines.[5] Organometallic reagents provided the first reports
that were often unexpected outcomes. These examples were followed
by reports of vinyliminium reactions with vinylogous amides and enols
to form an intermediate triene that underwent pericyclization to a
cyclic six-membered heterocyclic diene.[6]The development of [3 + 3]-cycloaddition transformations has
been
achieved through identification of suitable and compatible reactive
dipolar adducts and stable dipoles (Scheme 1). The reactive dipolar species is an energetic dipolar intermediate
that is generally catalytically formed in the reaction. The stable
dipole is any one of a spectrum of nitrones, ylides, and covalent
precursors to dipolar species. The two-step process occurs with the
reactive dipolar adduct reacting as an electrophile or as a nucleophile
to form the first covalent bond, and this association provides entropic
assistance for the construction of the second covalent bond and the
overall formal [3 + 3]-cycloaddition. The success of the overall process
is a function of the compatibility of the two reacting dipolar partners
and, of course, of the catalytic or stoichiometric species employed
for the generation of the reactive dipolar adduct.
Scheme 1
Reactive Dipolar
Species and Stable Dipoles Form [3 + 3]-Cycloaddition
Products
Organocatalysis
There are two basic catalytic pathways with which to enter [3 +
3]-cycloaddition transformations: organocatalysis and transition metal
catalysis. Organocatalysis is commonly based on vinyliminium ion chemistry,[7] but phosphine catalysis might also be compatible,[8] and N-heterocycliccarbene catalysis
has been reported.[9] Its formal relationship
to [3 + 3]-cycloaddition is similar to that of the linkage of the
Robinson annulation reaction to [4 + 2]-cycloaddition in that the
overall process is stepwise, rather than concerted, involving addition
and condensation. The vinyliminium ion is activated for either 1,2-
or 1,4-addition (Scheme 2), and both processes
are observed.
Scheme 2
Vinyliminium Ions Initiate [3 + 3]-Cycloaddition by
1,2- or 1,4-Addition
Extensive contributions in the organocatalysis approach
have been
made by Richard P. Hsung[7] who initially
recognized[6] that reactions of vinyliminium
ions with vinylogous amides formed an initial adduct that, upon elimination
of amine, formed a conjugated hexatriene, which could undergo pericyclic
rearrangement (tandem Knoevenagel/6π-electron electrocyclic
ring closure pathway) to azacyclohexadienes (Scheme 3),[10] which are precursors for piperidinyl
heterocycles or to 2H-pyrans.[11] In this pathway, a heteroatom (Z, mainly nitrogen, but
also oxygen) activates 1,2-addition onto the iminium ion, and reversible
β-elimination occurs to form a heterotriene that undergoes pericyclic
rearrangement to form the formal [3 + 3]-cycloaddition product. Often
the enamine (Z = NR2) and the iminium ion are derived from
the same unsaturated aldehyde or ketone. An extensive array of heterocyclic
compounds have resulted from this methodology.[7] However, enantiocontrol for these intermolecular reactions is not
straightforward due to the azatriene intermediate, although a high
level of stereocontrol has been achieved with the use of chiral auxiliaries
(Z = N-chiral auxiliary),[12] and structural modifications in the reactant pairs have formed [3
+ 3]-cycloaddition products by mechanisms that circumvent the azatriene
intermediate and provide high enantioselectivity.[13]
Scheme 3
Mechanistic Outline for Intermolecular Formal [3 +
3]-Cycloaddition
of Enamines with Vinyliminium Salts
Although the structural outcome is similar to that for
intermolecular
reactions, the operative mechanism for intramolecular [3 + 3]-cycloaddition
allows greater opportunity for enantiocontrol (Scheme 4).[10f] In this case, initial 1,4-addition
occurs, followed by tautomerization and 1,2-addition with elimination
of the amine and tautomerization to the final product. Asymmetric
induction is controlled by the amine catalyst used to form the iminium
ion reactant. Although enamine reactants are most common for this
transformation, 1,3-diketones or ketoesters are suitable alternatives
that form oxygen heterocycles via intramolecular reaction with an
imbedded dicarbonyl[10c] or by initial Michael
addition that forms the reactant for intramolecular cycloaddition.[11c,14]
Scheme 4
Mechanistic Outline for Intramolecular Formal [3 + 3]-Cycloaddition
of Enamines with Vinyliminium Salts
Procedural variants in the sequence of reactions of Schemes 3 and 4,[11] including sequences that employ bulky amine catalysts to
effect intermolecular vinylogous Michael addition to the vinyliminium
ion,[13a,13b,13d] have been
reported. In addition, diverse enamine methods have been developed
that utilize enones and enals in combination with acid catalysts for
the construction of six-membered ring heterocyclic[7,15] and
carbocyclic[13d,16] compounds. Reactants other than
enamines that include acrylic acid chlorides and esters[17] as well as allenes[18] and phosphorus ylides[19] and dipolar species
that include aziridines[15a,20] and cyclopropanes[21] have been employed. Several reviews have summarized
the diversity of heterocyclic compounds that can be prepared through
the organocatalytic approach.[7]
Transition Metal Catalysis: Prior Investigations
Until recently,
scarce attention has been given to transition metal-catalyzed
[3 + 3]-cycloaddition reactions other than for their uses as Lewis
acids. Following reports by Huang and co-workers with enolates and
presumed palladium acetate generated trimethylenemethane,[3] Hayashi reported a palladium-catalyzed [3 + 3]-cycloaddition
between azomethine imines, as well as between an N,α-diarylnitrone and the reactive dipoletrimethylenemethane,
generated from [2-(acetoxymethyl)-2-propenyl]trimethylsilane, that
occurred in high yield.[22] In these cases,
the catalyst converts the precursor molecule to a reactive dipolar
molecule that contains the ligated transition metal, which then undergoes
stepwise dipolar cycloaddition to a stable dipole (Scheme 5). Although both electrophilic and nucleophilic
adducts are conceivable in transition metal-catalyzed reactions, only
the electrophilic adducts have been reported thus far.
Scheme 5
Transition
Metal Catalyzed Generation of the Reactive Dipolar Species
for [3 + 3]-Cycloaddition
Azomethine imines have proven to be a model stable dipole
template
for [3 + 3]-cycloaddition reactions with trimethylenemethane (Pd catalysis)[22] and propargyl esters (Au catalysis),[23] but other pairings have been limited (Scheme 6).[24] However, the variety
of proposed metal-associated reactive dipoles indicates a high potential
for such reactive intermediates. Although [3 + 3]-cycloaddition between
1,3-dipoles (C-heteroarylimines) and electrophilic
Fischer vinylcarbene complexes was reported 15 years ago,[25] it was Toste and co-workers’ presentation
of a gold-catalyzed reaction between a propargyl benzoate, involving
a putative vinylcarbene intermediate, and azomethine imines that suggested
the catalytic potential of vinylcarbene species for [3 + 3]-cycloaddition
processes.[23]
Scheme 6
Transition Metal
Catalyzed [3 + 3]-Cycloaddition Reactions
[3 + 3]-Cycloaddition Reactions of Enoldiazoacetates
Vinylcarbenes are potential reactive dipolar adducts for [3 + 3]-cycloaddition
reactions. Their resonance contributing structures (Scheme 7) suggest a variety of transformations if the dipolar
characteristics of the vinylcarbene can dominate over both its carbenic
character[26] and activation of the conjugated
double bond for [2,3]-cycloaddition.[27] The
transition metal and its associated ligands establish the nucleophilic
or electrophilic character of the carbene, and R1 enhances
or diminishes its reactivity.[28] Vinyl substituents
R3 and R4 enhance the electrophilic or nucleophilic
character of the carbon to which they are attached, but it is the
substituent R2 that determines the viability of the [3
+ 3]-cycloaddition transformation. Conceivably, both electrophilic
and nucleophilic vinylcarbenes could undergo [3 + 3]-cycloaddition
transformations, but this transformation has only been observed with
electrophilic metal vinylcarbenes.
Scheme 7
Resonance Contributing Structures
of Vinylcarbenes Suggest Reaction
Outcome
[3 +
3]-Cycloaddition Reactions of Enoldiazoacetates
with Nitrones
The Doyle group recently reported an enantioselective
formal [3 + 3]-cycloaddition of nitrones with an enoldiazoacetate.[28] 3,6-Dihydro-1,2-oxazines are produced in high
yields with high enantiocontrol when catalyzed by Hashimoto’s
chiral dirhodium carboxylate Rh2(S-PTA)4 catalyst. The reaction mechanism is proposed to involve initial
dirhodium-catalyzed dinitrogen extrusion to form an intermediate metalenolcarbene. Nucleophilic attack by the nitrone at the vinylogous
position of the vinylcarbene followed by intramolecular iminium ion
electrophilic addition to the catalyst-activated vinyl ether functional
group with elimination of the dirhodium catalyst, in a stepwise or
concerted fashion (Nuc-Elec [3 + 3]-cycloaddition),
completes the transformation. During the cyclization, iminium ion
addition is facilitated by stabilization from the TBSO group, and
subsequent or concurrent release of the catalyst is a favorable process
(Scheme 8). Enantioselection is conferred by
the catalyst in the ring-closing step, and the degree of enantiocontrol
obtained is consistent with the intimate involvement of the catalyst
in the transition state for cyclization. Although [3 + 2]-cycloaddition
between nitrones and α,β-unsaturated carbonyl compounds,[29] and even with rhodium vinylcarbenes,[27] is known, this transformation did not occur
in the reactions of enoldiazoacetate 12 with nitrones.
Scheme 8
[3 + 3]-Cycloaddition of Catalytically Generated Enolcarbene Reactive
Dipoles with Nitrones
The reaction product is dependent on the transition metal
compound
used with enoldiazoacetate 12 and nitrones.[28,30] Dirhodium(II) catalysts direct the overall process solely to the
product from [3 + 3]-cycloaddition, whereas Lewis acids promote Mannich-type
addition as the sole outcome from both copper(I) and other Lewis acid
catalysts (Scheme 9).[31] Both transformations occur from the vinylogous position of the diazo
compound or its corresponding carbene. Reaction selectivity for diazonium
ion generation is determined, at least in part, by the ability of
the catalyst to coordinate with the stable dipole, which enhances
its electrophilic character relative to undergoing addition at the
diazo carbon, which results in metal carbene formation following extrusion
of dinitrogen. The absence of dinitrogen extrusion from reaction between
enoldiazoacetate and copper(I) catalysts is particularly surprising
in view of the known ability of copper(I) catalysts to generate carbene
intermediates with diazo compounds.[26]
Scheme 9
Competition between Metal Carbene Formation and Diazonium Ion Generation
Stepwise
[3 + 3]-Cycloaddition Reactions of
Enoldiazoacetates with Hydrazones and Azomethine Imines
The
stepwise nature of the cycloaddition process is evident in the reaction
of hydrazones with enoldiazoacetate 16.[32] A vinylogous 1,4-N–H insertion/Mannich addition
cascade reaction gives 1,2,3,6-tetrahydropyridazines 18 in good overall yields with up to 97% ee (Scheme 10). In this formal [3 + 3]-cycloaddition transformation excellent
enantioselectivities and high diastereoselectivities are controlled
by the chiral dirhodium(II) catalyst and Lewis acid, respectively.
The transformation is initiated by Rh(II)-catalyzed dinitrogen extrusion
followed by a previously unprecedented vinylogous 1,4-N–H insertion
into the hydrazone’s N–H bond, presumably through ammonium
ylide intramolecular proton transfer with elimination of the catalyst.[33] The dirhodium catalyst is insufficiently Lewis
acidic to activate the imine for ring closure; instead, Lewis acid
promoted Mannich addition of 19 smoothly generates 1,2,3,6-tetrahydropyridazines 18 with high diastereocontrol (Scheme 11).
Scheme 10
Two-Step, One-Pot [3 + 3]-Cycloaddition of Catalytically Generated
Enolcarbene Reactive Dipoles with Hydrazones
Scheme 11
Mechanism of [3 + 3]-Cycloaddition by 1,4-N–H Insertion
of
Catalytically Generated Enolcarbene Reactive Dipoles with Hydrazones
Followed by Lewis Acid Directed Ring Closure
Guided by the formal [3 + 3]-cycloaddition reactions of
enoldiazoacetates
with nitrones and hydrazones, various 1,3-dipoles have been examined
for their compatibility. With azomethine imines, a highly regio- and
diastereoselective [3 + 3]-annulation reaction with enoldiazoacetates
gives bicyclic pyrazolidinone derivatives 20 when R1 is an alkyl, an aryl, or a vinyl group (Scheme 12).[34] However, when R1 is hydrogen, N–N-cleavage of the azomethine imine
occurs, and imine derivative 21 is obtained. The two
different outcomes in this reaction occur by nucleophilic attack of
the azomethine imine on the metal enolcarbene at the vinylogous position
(R1 = alkyl, aryl, or vinyl) or at the metal carbene center
(R1 = H), but the precise cause is unknown.
Scheme 12
[3 +
3]-Cycloaddition of Catalytically Generated Enolcarbene Reactive
Dipoles with Azomethine Imines
Dearomatization in [3 + 3]-Cycloaddition Reactions
of Enoldiazoacetates
Unlike in reactions of nitrones, hydrazones,
and azomethine ylides with enolcarbene intermediates, reactions with N-iminopyridinium ylides afford a barrier to cycloaddition
due to dearomatization, and for this reason, it is perhaps not surprising
that [3 + 3]-cycloaddition reactions with N-iminopyridiniumylides have only been reported in limited cases with 1,1-cyclopropane
diesters.[35] However, application of N-acyliminopyridinium ylides (22) as stable
dipoles in reactions with enoldiazoacetates catalyzed by dirhodium(II)
catalysts gave the [3 + 3]-cycloaddition product in high isolated
yields and with exceptional enantiocontrol when catalyzed by Rh2(S-PTTL)4 and Rh2(S-PTAD)4 (Scheme 13).[36] In this transformation, steric effects have
an important influence on the control of selectivity since dirhodium
catalysts with bulky ligands [Rh2(S-PTTL)4 instead of Rh2(S-PTA)4] give dramatic improvements in selectivity control. Either pyridine,
quinoline, or isoquinoline derived ylides work very well in these
[3 + 3]-cycloaddition reactions, giving up to 98% ee in reactions
with enoldiazoacetates.
Scheme 13
[3 + 3]-Cycloaddition of Catalytically
Generated Enolcarbene Reactive
Dipoles with N-Iminopyridinium Ylides
The stable dipoles described thus far were not
sufficiently basic
to cause inhibition of dirhodium catalysts toward metal carbene formation.
These dirhodium(II) compounds are mild Lewis acids that coordinate
with Lewis bases,[37] sometimes causing diminished
reactivity toward diazo compounds. Isoquinolinium/pyridinium methylides
are readily accessible nucleophiles of variable base strengths[38] that readily undergo [2 + 3]-cycloaddition reactions.[39] Testing the limits of stable dipoles in their
reactions with metalloenolcarbenes, isoquinolinium/pyridinium methylides
were treated with enoldiazoacetate 12 in the presence
of dirhodium catalyst without obvious catalyst inhibition, and instead
of 1,3-dipolar cyclization through [2 + 3]-cycloaddition, [3 + 3]-cycloaddition
readily occurred to give substituted quinolizidines 26 in high yield and high enantioselectivity when the reaction was
catalyzed by Rh2(S-PTIL)4 (Scheme 14).[40] The reaction outcome
was solvent, catalyst, and temperature dependent with a competing
process that formed an apparent product from [3 + 2]-cycloaddition
of 24 to donor–acceptor cyclopropene 25.
Scheme 14
[3 + 3]-Cycloaddition of Catalytically Generated Enolcarbene
Reactive
Dipoles with Isoquinolinium/Pyridinium Methylides
Donor–Acceptor Cyclopropenes
as Reactive
Enolcarbene Sources in [3 + 3]-Cycloaddition Reactions
During
investigations with isoquinolinium/pyridinium methylides, a unique
equilibrium was revealed by the competitive formation of [2 + 3]-cycloaddition
product 27 in amounts that varied with the amount of
catalyst employed. Coordination of Lewis basic methylides to dirhodium(II)
was established, and this association was proposed to prompt rearrangement
of the enolcarbene bound to dirhodium(II) to produce donor–acceptor
cyclopropene 25. However, independently formed donor–acceptor
cyclopropene 25 was also demonstrated to be a precursor
of the same metal carbene intermediate that formed 26 (identical enantioselectivities), and via this process, the donor–acceptor
cyclopropene is in equilibrium with the dirhodium-bound enolcarbene
(Scheme 15). The reaction pathways in this
system involve enantioselective [3 + 3]-cycloaddition from the dirhodium-bound
enolcarbene and uncatalyzed diastereoselective [3 + 2]-cycloaddition
of cyclopropene 25 with isoquinolinium or pyridinium
methylides.
Scheme 15
Involvement of Donor–Acceptor Cyclopropenes
in Cycloaddition
Reactions of Catalytically Generated Enolcarbene Reactive Dipoles
with Isoquinolinium/Pyridinium Methylides
Variation in the silyl ether and carboxylate ester groups
of enoldiazoacetates
maintains the [3 + 3]-cycloaddition process that occurs through a
metal carbene pathway in dirhodium(II)-catalyzed reactions, but changing
the substituents at the vinylogous position can inhibit this cycloaddition
pathway. For example, when γ-phenyl enoldiazoacetate 28 underwent dinitrogen extrusion with rhodium(II) acetate in the presence
of diphenylnitrone 13, there was no observable product
from [3 + 3]-cycloaddition even though O–H insertion into benzyl
alcohol readily occurred. However, catalysis by Cu(SbF6)2 and AgSbF6 gave the [3 + 3]-cycloaddition
product in high yield within 5 min under the same conditions (Scheme 16).[41] The reaction pathway
with Cu(II) and Ag(I) changed from that with dirhodium(II) and is
consistent with one involving Lewis acid addition to the diazo carbon
to form a diazonium ion intermediate (Scheme 16).
Scheme 16
Dipolar Intermediates Catalytically Generated for Metal Carbene
or
Lewis Acid Adduct [3 + 3]-Cycloaddition Reactions
In an effort to effect enantiocontrolled cycloaddition
of γ-phenyl-enoldiazoacetate 28a with nitrones,
AgSbF6/(S)-BuBox catalyst was found to be superior to
all other Lewis acid/ligand combinations used, giving the [3 + 3]-cycloaddition
product in 92% yield but with only 61% ee under conditions limited
by the temperature necessary to initiate the reaction (eq 1, −30 °C). However, with the corresponding
donor–acceptor cyclopropene generated in situ from γ-phenyl-enoldiazoacetate 28 through catalysis
by rhodium(II) acetate, formation of the [3 + 3]-cycloaddition product
could be optimized (Scheme 17) to 93% yield
with 90% ee (at −78 °C with the Bu ester 28b).[42] Here
the question arises of whether the silver(I)-catalyzed reaction is
the result of a process that occurs through an organometallic intermediate
formed by electrophilic addition of Ag(I) to the carbon–carbon
double bond of the donor–acceptor cyclopropene or, as is suggested
from results with dirhodium(II) catalysts on the cyclopropene analogues
without the phenyl substituent, through a silver-enolcarbene intermediate.[43] Enantioselectivities from the reactions of 28 and the corresponding donor–acceptor cyclopropene
(30) with AgSbF6/(S)-BuBox are not exactly the same (61% ee vs
67% ee), so both intermediates must be considered.
Scheme 17
[3 + 3]-Cycloaddition of Silver(I)-generated Reactive Dipoles
from
Donor–Acceptor Cyclopropenes with Nitrones
Summary and Outlook
The success of enoldiazoacetates in achieving highly selective
[3 + 3]-cycloaddition transformations can be attributed to several
factors, among which are steric influences around the metal carbenic
center that inhibit attachment of nucleophiles at the carbonic center
lending preference to vinylogous attachment. Another advantage is
the electrophilic character imparted to the vinylogous position by
the ligated transition metal of the metal carbene (Scheme 7). However, it is the electron-donating oxygen of
the silyl ether that enhances the viability of electrophilic ring
closure to the metal-bound carbon (Schemes 8, 11, 13, and 16) and inhibits the alternative [3 + 2]-cycloaddition.
The extent to which other electron-donating substituents will favor
[3 + 3]-cycloaddition is yet to be determined, but replacement of
the silyl ether by hydrogen (but not phenyl) in reactions with an
isoquinolinium methylide (Scheme 14) that produces
the [3 + 3]-cycloaddition product exclusively in modest yield suggests
broad application.[40] That donor–acceptor
cyclopropenes are suitable precursors to the same enolcarbene intermediates
formed by catalytic dinitrogen extrusion of enoldiazoacetates provides
an alternative entry to [3 + 3]-cycloaddition reactions that is just
now being investigated; however, just as with enoldiazoacetates and
as was found with organocatalysis, multiple mechanistic pathways to
the same reaction products are possible, and alternatives to dirhodium(II)
catalysts may be viable.In conclusion, catalytically generated
enolcarbenes are effective
reactive dipolar species for reactions with stable dipolar compounds,
and they are key to the development of [3 + 3]-cycloaddition reactions
as a complementary strategy to alternative [4 + 2]-cycloaddition for
the synthesis of heterocyclic compounds. In particular, metallo-enolcarbenes
that are generated catalytically from conveniently prepared and highly
stable enoldiazoacetates or from donor–acceptor cyclopropenes
are highly effective dipolar adducts that give six-membered ring heterocyclic
products, not only with nitrones,[28,41,42] hydrazones,[32] and ylide
derivatives,[36,40] but also with nitrile oxides,[44] oximes,[45] imines,[46] and donor–acceptor substituted hydrazones[47] (Scheme 18). With this
methodology, one or more heteroatoms can be introduced into the six-membered
ring in high yields and with high levels of stereocontrol obtained
through the use of asymmetric catalysts.[48] Further applications of these enoldiazoacetates are worth pursuing
with new catalyst development for the stereoselective synthesis of
functionalized heterocyclic compounds.
Scheme 18
Heterocyclic Syntheses
from Cycloaddition Reactions with Enoldiazoacetates
Authors: Yongming Deng; Lynée A Massey; Yeray A Rodriguez Núñez; Hadi Arman; Michael P Doyle Journal: Angew Chem Int Ed Engl Date: 2017-08-24 Impact factor: 15.336