Phosphorus-based organocatalysis encompasses several subfields that have undergone rapid growth in recent years. This Outlook gives an overview of its various aspects. In particular, we highlight key advances in three topics: nucleophilic phosphine catalysis, organophosphorus catalysis to bypass phosphine oxide waste, and organophosphorus compound-mediated single electron transfer processes. We briefly summarize five additional topics: chiral phosphoric acid catalysis, phosphine oxide Lewis base catalysis, iminophosphorane super base catalysis, phosphonium salt phase transfer catalysis, and frustrated Lewis pair catalysis. Although it is not catalytic in nature, we also discuss novel discoveries that are emerging in phosphorus(V) ligand coupling. We conclude with some ideas about the future of organophosphorus catalysis.
class="Chemical">Phosphorus-based organocatalysis enclass="Chemical">pan class="Chemical">compasses several subfields that have undergone rapid growth in recent years. This Outlook gives an overview of its various aspects. In particular, we highlight key advances in three topics: nucleophilic phosphine catalysis, organophosphorus catalysis to bypass phosphine oxide waste, and organophosphoruscompound-mediated single electron transfer processes. We briefly summarize five additional topics: chiral phosphoric acid catalysis, phosphine oxide Lewis base catalysis, iminophosphorane super base catalysis, phosphonium salt phase transfer catalysis, and frustrated Lewis pair catalysis. Although it is not catalytic in nature, we also discuss novel discoveries that are emerging in phosphorus(V) ligand coupling. We conclude with some ideas about the future of organophosphorus catalysis.
The study and application of organicclass="Chemical">phosphorusclass="Chemical">pan class="Chemical">compounds in organic
synthesis have a long history that dates back to the Middle Ages.[1] Scientific investigations into these types of
compounds, especially from a synthetic perspective, have unveiled
the rich chemistry of organophosphoruscompounds. In this realm, Michaelis
and Arbusov, as well as other chemists, were the pioneers who laid
the foundations of what we know as phosphorus chemistry today.[1] One significant feature of organic phosphoruscompounds is that they can be used as ligands, catalysts, or reagents
to facilitate various organic reactions.[2,3] When asked
about the roles of phosphoruscompounds in organic synthesis, however,
a majority of organic chemists might offer stereotypical answers:
ligands in transition-metal-promoted reactions, reagents for Wittig
and Mitsunobu reactions, and organocatalysts for reactions that corresponding
nitrogencounterparts also facilitate [e.g., the Morita–Baylis–Hillman
(MBH) reaction]. In truth, phosphorus-based organocatalysis is a unique,
multifaceted field, with many promising research opportunities and
applications. For example, chiral phosphoric acid catalysis[4] and phosphine oxide Lewis base catalysis[5] play important roles in asymmetric organocatalysis.
Frustrated Lewis pairs (FLPs), combinations of Lewis acids and Lewis
bases, are important species for the activation of small molecules.[6] As a green sustainable tool in organic synthesis,
phosphonium salt-based phase transfer catalysis is garnering increasing
attention.[7] Bifunctional phosphonium catalysts
have especially great potential in asymmetric synthesis. Chiral iminophosphoranesuper base catalysts have demonstrated their usefulness in the deprotonation
of high-pKa pronucleophiles for enantioselective
transformations.[8] An abundance of structural
and stereochemical diversity among readily available phosphoruscompounds
remains pivotal for developing novel asymmetric transformations. Notably,
nucleophilic phosphine catalysis is being investigated vigorously,
with discoveries of novel reaction patterns and asymmetric transformations.[3] Although classic transformations that are driven
by the formation of phosphine oxides (e.g., Wittigolefination, Mitsunobu,
and Appel reactions) are highly popular among synthetic chemists,
these reactions have several drawbacks [e.g., the requirement of stoichiometric
phosphorus reagents, the use of external oxidants (e.g., azo compounds
for Mitsunobu reactions), and the generation of phosphine oxide waste]
that limit their use in both academia and industry. Developing catalytic
versions of these reactions—in particular to bypass phosphineoxide waste—is highly desirable.[3] This area has been studied intensely over the past decade. In addition
to those phosphorus-mediated reactions that operate under mechanisms
involving electron pair transfer, phosphoruscompound-mediated single-electron
transfer (SET) processes are a promising, yet undeveloped, area. Because
of the rapid development and growing interest in photochemistry and
electrochemistry in recent years, studies of organophosphoruscompound-mediated
SET processes are gaining momentum.[9]
In the short span of two years since our class="Chemical">comclass="Chemical">prehensive review on
class="Chemical">pan class="Chemical">organophosphorus catalysis in 2018,[3b] significant
progress has been made in the field. In this Outlook, we highlight
the key features of three such topics: nucleophilic phosphine catalysis,
organophosphorus catalysis to bypass phosphine oxide waste, and organophosphoruscompound-mediated SET processes. We also provide an overview of several
other important topics related to phosphorus-based catalysis in the Miscellaneous Catalyses section. We do not go into
too much detail regarding these miscellaneous topics but instead draw
the reader’s attention to several comprehensive reviews. Moreover,
interesting and useful chemistry is emerging in phosphorus-based reactions.
For example, phosphorus(V) ligand coupling, although not catalytic
in phosphorus, is becoming a method complementary to transition-metal-catalyzed
coupling for the functionalization of heterocycles; we discuss these
processes briefly. Although this Outlook is not intended to provide
a comprehensive review, we are aiming to give an overview of phosphorus-based
catalysis, point out the key developments in the past two years, and
proffer our perspective about the future of phosphorus-based catalysis,
hoping to inspire researchers to consider the rich chemistry of organophosphoruscompounds in their own endeavors.
Nucleophilic
Phosphine Catalysis
Nucleophilicclass="Chemical">phosphine catalysis tyclass="Chemical">pically
oclass="Chemical">perates through initial
adclass="Chemical">pan class="Disease">dition of a tertiary phosphine to an electron-deficient multiple
bond, generating a reactive zwitterionic species. Various substrates,
including alkenes, allenes, alkynes, and MBH alcohol derivatives (MBHADs),
can be involved in this type of activation. The generated reactive
zwitterionic intermediates can undergo further reactions with a variety
of nucleophiles and electrophiles, or a combination of both. Nucleophilic
phosphines also play a special role in the activation of C=O,
N=N, strained C—C, or strained C—X (e.g., C—O,
C—N) bonds. The broad substrate scope, mild reaction conditions,
and diverse reaction modes have established nucleophilic phosphine
catalysis as a powerful tool in organic synthesis. Since our group’s
comprehensive review of phosphine organocatalysis in 2018,[3b] more than 80 new reports of nucleophilic phosphine
catalysis have been published.[10] A survey
of the progress in nucleophilic phosphine catalysis for the past two
years has revealed some interesting features. The main thrusts of
the research continue to be the discovery of new patterns of annulation,
asymmetric variants of previously reported reactions, and investigations
into known reactions using substrates that are much more elaborate
than those reported originally. In this Outlook, we classify the reactions
of nucleophilic phosphine catalysis based on the nature of their electrophiles
with different types of activated multiple bonds—specifically,
alkenes, allenes, alkynes, and MBHADs. We further categorize these
reactions in terms of the second reaction partner—namely, nucleophiles,
electrophiles, and their combinations. In this section, we discuss
some recent representative examples of nucleophilic phosphine catalysis
to illustrate its unique features and rich chemistry.
Nucleophilic Phosphine Catalysis of Alkenes
The nucleophilicclass="Chemical">phosphine catalysis of class="Chemical">pan class="Chemical">alkenes involves the addition
of phosphine to electron-deficient alkenes 2–1 to form Horner’s zwitterionic adduct 2–2, which has multifarious modes of reaction
depending on its reaction partners (Scheme ). As such, reactions of the zwitterion 2–2 with a pronucleophile (NuH), an electrophile
(E), or a hybrid of both can lead to various synthetically useful
scaffolds. Recent reports on the reactions of alkenes with pronucleophiles
have been limited to the phosphine-catalyzed Michael addition, in
which pronucleophiles, such as phosphine oxides and TMSCN, have been
introduced to produce β-phosphoryl- and β-cyano-carbonyls,
respectively.[11] Among the reactions of
the zwitterionic adduct with electrophiles, the MBH and Rauhut–Currier
(RC) reaction are regularly mentioned as representative classical
examples. The broad scope and general applicability of these two reactions
for the construction of acyclic and carbo- and heterocyclic compounds
have undoubtedly contributed to the significant attention that they
have garnered from synthetic chemists. Recent progress in these two
reactions has focused mainly on using elaborated alkenes to construct
complex polycyclic structures in an enantioselective manner.[12,13] In an elegant example, in 2019, Vicario and co-workers reported
a catalytic enantioselective transannular MBH reaction to afford various
fused bicycliccompounds containing stereogenic tertiary alcohol moieties
in excellent yields and enantioselectivities (Scheme ).[13a] In their
proposed model 2–6 for stereoinduction,
hydrogen bonding interactions between the NH functional groups of
the bifunctional phosphine catalyst and the electrophilic ketone unit
of the substrate were crucial to impart the high enantioselectivity.
To further demonstrate its utility, they adopted this new method in
the first asymmetric synthesis of the sesquiterpene natural product
γ-gurjunene. Specifically, when the racemic ketone 2–7 was subjected to the conditions, kinetic resolution
occurred to afford the desired bicyclic adduct 2–8 in 42% yield and 90% ee, paving the way for the total synthesis
of γ-gurjunene. In another recent example of reactions between
the zwitterionic adduct and electrophiles, Lin and co-workers reported
the direct β-acylation of 4-arylidene pyrazolones and 5-arylidene
thiazolones with acyl chlorides via phosphine catalysis.[13b] When an electrophile–nucleophile is
reacted with an alkene, annulation usually occurs. For example, Guo
and co-workers reported an efficient annulation between γ-sulfonamido-enones
and sulfamate-derived cyclic imines to furnish multicyclic imidazolidine
derivatives.[14]
Scheme 1
Nucleophilic Phosphine
Catalysis of Alkenes
Scheme 2
Asymmetric Transannular
MBH Reaction
Tertiary class="Chemical">phosphines
add to electroclass="Chemical">philes even in the class="Chemical">presence of
electron-deficient class="Chemical">pan class="Chemical">alkenes. A recent paper by Zhang and co-workers
unveiled a phosphine-catalyzed vicinaldifunctionalization of α,β-enones
with TMSN3 (Scheme ).[15a] In this unique transformation,
the reaction was initiated by the addition of 1,4-bis(diphenylphosphino)butane
(dppb) to TMSN3 to form a phosphazide zwitterion 2–12. Michael addition of this intermediate
followed by intramolecular cyclization generated 2–13, whose decomposition formed the β-amino α-diazo
carbonyl compounds 2–11. To slow
down the decomposition of the phosphazide zwitterionic intermediate 2–12 to the iminophosphorane, the bulky
TMS group was intentionally selected. A bifunctional chiral phosphine
catalyst 2–14 promoted the asymmetric
version of this novel transformation, affording moderate enantioselectivities.
In a related study, the same group reported a phosphine-catalyzed
[3 + 2] cycloaddition of α-diazoacetates and β-trifluoromethyl
enones to access 4-(trifluoromethyl)pyrazolines.[15b]
Scheme 3
Phosphine-Catalyzed Difunctionalization of α,β-Enones
Nucleophilic Phosphine
Catalysis of Allenes
class="Chemical">Phosphine catalysis of class="Chemical">pan class="Chemical">allenes is the
most active area of nucleophilic
catalysis, due to the diverse modes of reactivity displayed by allenes.[3b] In particular, highly functionalized allenes
that deviate from the generic versions used in classic reactions often
result in unforeseen reactivity patterns. Novel reactions continue
to be invented, often accompanied by their asymmetric variants mediated
by chiral phosphines, reflective of the potpourri of quality chiral
phosphines made available in the past decade, specifically for nucleophilic
organocatalysis. The phosphine catalysis of allenes reported in the
past two years is presented here in three sections, categorized based
on the nature of the second starting material: nucleophile, electrophile,
or electrophile–nucleophile.
Nucleophilic
Phosphine Catalysis of Allenes
with Nucleophiles
Reactions of Allenes with Nucleophiles
Umpoclass="Chemical">lung adclass="Chemical">pan class="Disease">dition
is the most prevalent mode of reaction between an electron-deficient
allene and a nucleophile in the presence of a tertiary phosphine catalyst.
Among the umpolung additions of allenes, β′- and γ-umpolung
additions are the most commonly observed reaction modes (Scheme ).[3b] For allenoates lacking an alkyl substituent at the α-position,
γ-umpolung addition may occur; α-alkyl-substituted allenoates
undergo β′-umpolung addition. Less common is the Michael
addition onto the β-carbon of allenoates, as has been demonstrated
with α-fluoro-β-ketoamide nucleophiles.[16] Enantioselective γ-umpolung additions of allenes
with nitrogen-centered nucleophiles have continued to be studied.[17] The Huang group reported a novel phosphine-catalyzed
ε-umpolung addition of vinyl allenoates that affords diene carboxylates
(Scheme ).[18] Depending on the number of hydrogen atoms on
the pronucleophile, either mono or double remote 1,7-addition products
can be obtained. Mechanistically, the ε-umpolung addition can
be viewed as a vinylogous γ-umpolung addition triggered by the
γ-vinyl substituent on the allenoate. Nucleophilic addition
of a phosphine catalyst to the γ-vinyl allenoate 2–15 would form the zwitterion 2–18. Subsequent deprotonation of the pronucleophile and nucleophilic
addition at the remote ε-carbon atom, followed by proton transfer
and β-elimination of the catalyst, would afford the product 2–16. Among the series of phosphines tested,
the bifunctional phosphine 2–17 was
the best agent. Huang et al. reasoned that the reaction rate was accelerated
as a result of additional interactions between the multifunctional
phosphine and the substrates.
Scheme 4
β′- and γ-Umpolung
Nucleophilic Addition
Scheme 5
ε-Umpolung Nucleophilic 1,7-Addition of Vinyl Allenoates
Reactions of Allenes with Dinucleophiles
and Trinucleophiles
While simple nucleophiles are known to
promote umpoclass="Chemical">lung adclass="Chemical">pan class="Disease">dition
of allenes, a distinctive [m + n] mode of annulation can occur when dinucleophiles or trinucleophiles
are introduced. Furthermore, when combined with multifunctionalized
allenes, these nucleophiles can result in several unique modes of
annulation. For example, phosphine-catalyzed [3 + 2] annulations of
the dinucleophile 2-sulfonamidomalonate with δ-acetoxy allenoates
have allowed access to pyrrolines;[19a] when
vinyl allenoate was reacted with trinucleophiles, complex tetracyclic
structures were constructed with the formation of two new rings.[19b] Along with those novel annulation processes,
new patterns of [4 + 1] annulation to form 5-membered carbo- or heterocycles
have been discovered.[10b,20] In an elegant example, Huang
and co-workers reported a novel phosphine-catalyzed [4 + 1] cyclization
of allenyl imides with methyl ketimines and enamines (Scheme ).[20a] This method provided efficient access to cyclopentenoyl enamines
and imines. The α-phosphonium vinyl methyl ketene 2–21 was proposed as the key 1,4-(bis)electrophile
intermediate, which reacted with the nucleophiles 2–22 and 2–24 to afford the
cyclopentenones. Notably, changing the traditional allenoates to allenyl
imides resulted in unforeseen reactivity patterns.
Scheme 6
[4 + 1] Cycloadditions
of Allenyl Imides with Methyl Ketimines and
Enamines
Nucleophilic
Phosphine Catalysis of Allenes
with Electrophiles
Unlike the adclass="Disease">dition reactions of nucleoclass="Chemical">philes
with class="Chemical">pan class="Chemical">allenes, the majority of reactions of allenes with electrophiles
forge two new bonds to afford cyclic scaffolds.[3b] A diverse range of electrophiles (e.g., alkenes, imines,
ketones/aldehydes, aziridines, and azomethine imines) can be employed,
rendering phosphine-catalyzed reaction of allenes with electrophiles
a powerful tool in organic synthesis.
Lu’s Allene–Alkene/Imine
[3 + 2] Annulation
While annulations of class="Chemical">allenes with class="Chemical">pan class="Chemical">ketones,
aziridines, azomethine
imines, troponones, and other reaction partners have continued to
be explored in this area,[21] Lu’s
[3 + 2] annulation[22] has been the most
thoroughly studied and adopted for a variety of synthetic applications.[23] Researchers in the field have continued to find
new ways to reinvent Lu’s allene–alkene/imine [3 + 2]
annulation. Whereas engagement of previously unreported alkenes remains
a dominant activity,[24−27] significant attention has also been paid to the preparation of enantioenrichedcyclopentenes and pyrrolidines.[26,27] In 2019, Lu and co-workers
reported a phosphine-catalyzed enantioselective [3 + 2] annulation
reaction between allenes and isoindigos, resulting in spirooxindoles
having two vicinal quaternary stereogenic centers (Scheme ).[26a] The threonine-derived bifunctional phosphine 2–29 catalyzed the [3 + 2] annulation to afford dimeric spirocyclic
bisindoline motifs in high yields and excellent enantiomeric excesses.
Notably, the product 2–28 was transformed
into the bisoxindole 2–30, which
acted as a common intermediate for the bisoxindoles 2–31 [a known precursor of the natural products
(−)-ditryptophenaline and (−)-WIN 64821] and 2–32 [a known precursor of (−)-chimonanthine
and (−)-folicanthine].
Scheme 7
[3 + 2] Annulation of Isoindigos with
Allenes
Several examples of class="Chemical">allene–class="Chemical">pan class="Chemical">nitroolefin
[3 + 2] annulations
have been reported.[24b,25,26] In particular, the engagement of 3-nitroindoles, 3-nitrobenzofurans,
and 3-nitrobenzothiophenes resulting in the dearomative functionalization
of electron-deficient aromatic rings demonstrates that phosphine organocatalysis
is potentially a platform that can be complementary to metal-catalyzed
dearomative reactions (Scheme ).[25,26] Two elegant examples have appeared
from the Zhang[26b] and Lu[26c] groups, who demonstrated the asymmetric dearomative functionalization
of electron-deficient allenoates 2–33 with 3-nitroindoles 2–34 catalyzed
by the bifunctional phosphines 2–36 and 2–37, respectively. Compared
with the preparation of enantioenrichedcyclopentenes, examples of
enantioselective allene–imine [3 + 2] annulation are rare.
To overcome this deficiency, Kwon and co-workers prepared the novel
P-stereogenic [2.2.1] bicyclic chiral phosphine CarvoPhos from the
natural product carvone[27a] and applied
it to the asymmetric syntheses of a bevy of pyrrolines, including
the antiarrhythmogenic small molecule efsevin.
Scheme 8
Dearomative Functionalization
through Phosphine Catalysis
Allene–Electrophile [4 + 2] Annulation
Along
with the class="Chemical">allene–class="Chemical">pan class="Chemical">alkene/imine [3 + 2] cyclization, Kwon’s
α-alkylallenoate–imine/alkene [4 + 2] annulation[28] is another classic example of phosphine-catalyzed
synthesis of cyclic compounds.[29] The main
development on this reaction since our comprehensive review in 2018[3b] has been in the enantioselective syntheses of
tetrahydropyridines and cyclohexenes.[30] While Kwon and co-workers reported an enantioselective synthesis
of guvacine derivatives through the allene–imine [4 + 2] annulation
of imines with α-methylallenoates mediated by [2.2.1] bicyclicexo-(p-anisyl)-HypPhos,[30a] Lu and co-workers disclosed a chiral bifunctional phosphine-catalyzed
asymmetric [4 + 2] reaction to access functionalized tetrahydropyridines.[30c] 3-Nitroindoles and 3-nitrobenzothiophenes are
suitable reaction partners in the α-alkylallenoate–alkene
[4 + 2] annulation, as reported by Guo and co-workers for the (S)-SITCP-catalyzed synthesis of dihydrocarbazoles.[30d]
In the class="Chemical">context of class="Chemical">pan class="Chemical">allene–electrophile
[4 + 2] annulation, δ-acetoxy allenoates have been good precursors
for the synthesis of benzene rings.[31] In
a striking example, Tong and co-workers found that the phosphoniumdiene intermediate formed by addition of the phosphine to δ-acetoxyallenoatescould undergo [4 + 2] annulation with 2-hydroxyquinone derivatives
to form naphthoquinones 2–40 (Scheme ).[31a] The efficient and stereoselective synthesis of axially
chiral naphthaquinone atropisomers was facilitated by the cooperative
effects of the tertiary phosphine and tertiary amine in the bifunctional
chiral phosphine 2–41, as displayed
in the intermediate 2–42. In another
example, Swamy and co-workers demonstrated that the PPh3-catalyzed reactions between δ-acetoxy allenoates and cyclic N-sulfonyl iminescould also generate functionalized benzenes.[31c]
Scheme 9
Synthesis of Aryl-Naphthaquinone Atropisomers
Allene–Electrophile Multicycle Annulation
When
class="Chemical">allenes and/or electroclass="Chemical">philes having adclass="Chemical">pan class="Disease">ditional functionalities (e.g.,
conjugated C=C bonds, a leaving group, or a pronucleophile)
are employed, never-before-seen reactions, especially multicycle annulations,
may ensue to give complex polycyclic structures. For example, when
γ-benzylallenoates were treated with (E)-2-(3-arylallylidene)malononitriles
in the presence of a phosphine catalyst, multicycle annulation occurred
to give bicyclic[3,3,0]octene derivatives.[24a] In another example, a C2-symmetric chiral phosphine,
NUSIOC-Phos, promoted the reactions between pyrrolidine-2,3-diones
and γ-substituted allenoates, resulting in a domino process
to form γ-lactams containing five contiguous stereogenic centers.[32a] Furthermore, Huang and co-workers reported
a sequential annulation of δ-sulfonamidoallenoates 2–43 and dienes 2–44 to access highly functionalized hydroisoquinolines 2–45 (Scheme ).[32b] The δ-sulfonamidoallenoates
acted as novel five-atom units in the sequential annulations with
the involvement of an intermediate 2–46 with three nucleophilic sites.
Scheme 10
Domino Annulation of δ-Sulfonamidoallenoates
to Access Hydroisoquinoline
Derivatives
δ-Umpolung Addition
Given that most reactions
between class="Chemical">allenes and electroclass="Chemical">philes have been cyclizations, class="Chemical">pan class="Chemical">Xu’s
report of a catalytic asymmetric dienylation of para-quinone methides (p-QM) 2–48 with γ-benzylallenoate 2–47 through δ-umpolung addition is noteworthy (Scheme ).[33] Various optically pure diarylmethanes 2–49 were obtained in moderate to good yields through this method.
They proposed that the phosphonium dienolate intermediate 2–51 underwent proton transfer to afford the vinylogous
ylide 2–52. Subsequent 1,6-conjugate
addition to the p-QM 2–48, followed by a series of proton transfers and β-elimination
of the catalyst 2–50, afforded the
product 2–49.
Scheme 11
δ-Umpolung
Addition
Nucleophilic
Phosphine Catalysis of Allenes
with Electrophile–Nucleophiles
When class="Chemical">allenes react
with a molecule class="Chemical">pan class="Chemical">containing both electrophilic and nucleophilic groups
in the presence of catalytic phosphine, annulations typically occur.
Representative examples of nucleophile–electrophiles include
salicylaldehyde and salicylimine. Recently, Zhou and co-workers introduced
an additional pronucleophile, malonate, into the salicylimine structure
to give the dinucleophile–electrophile 2-[(2-hydroxybenzylidene)amino]
malonate.[34a,34b] When this dinucleophile–electrophile
was treated with benzyl β′-acetoxyallenoate, multicycle
annulation occurred to give polycyclic products.[34a,34b] Mohanan and co-workers found that the annulation of nucleophile–electrophile
phenacylmalononitriles and allenoates catalyzed by PBu3 produced functionalized cyclopentenes.[34c] Another phosphine-catalyzed reaction of nucleophile–electrophile
2-tosylaminochalcones and β′-acetoxyallenoates, reported
by the Huang lab, is also notable (Scheme ).[35a] The β′-acetoxyallenoate 2–54 was transformed into the enynoate
intermediate 2–58 through a substitution/elimination
sequence. Nucleophilic addition of the phosphine to 2–58 resulted in the zwitterion 2–59, which underwent 1,4-addition with 2–55 and subsequent intramolecular SN2displacement to form the tetrahydroquinoline 2–56. The Huang group extended this method to
the synthesis of functionalized chromans.[35b] This report was the first to describe addition across the α-
and β′-carbon atoms of α-substituted allenoates.
Scheme 12
Synthesis of 3-Ethynyl-Substituted Tetrahydroquinolines
Nucleophilic Phosphine
Catalysis of Alkynes
A variety of pronucleophiles, electrophiles,
and nucleophile–electrophiles
can be involved in the class="Chemical">phosphine catalysis of class="Chemical">pan class="Chemical">alkynes.[3b] Although phosphine catalysis with alkynes has
received less attention than that with allenes, reports of known reactions
with previously unexplored substrates, as well as discoveries of novel
reactions, have continued to appear. Among these transformations,
Michael addition and α-umpolung addition have been the two major
reaction pathways when nucleophiles are present (Scheme ). In one recent example,
Salin and co-workers reported a bishydrophosphorylation of electron-deficient
alkynes through nucleophilic phosphine catalysis.[36a] In another example, Zhang and co-workers demonstrated a
chiral bifunctional phosphine-promoted double Michael addition of
a tryptamine-derived oxindoledinucleophile onto ynones.[36b] Contemporarily, the Wang group reported the
efficient synthesis of (E)-2-nitromethylcinnamates
through phosphine catalysis of alkynes (Scheme ).[37a] In this
transformation, initial phosphine-catalyzed α-umpolung addition
of the alkyne 2–62 with the nitromethane 2–61 formed the 1,2-disubstituted alkene
intermediate 2–64. Subsequent phosphine-promoted
1,3-rearrangement of the nitro group generated the more stable trisubstitutedalkene 2–63.
Scheme 13
Michael Addition
and α-Umpolung Addition in the Phosphine Catalysis
of Alkynes
Scheme 14
Synthesis of (E)-2-Nitromethylcinnamates
through
Phosphine Catalysis of Alkynes
Anticarboboration, silaboration, and class="Disease">diboration of class="Chemical">pan class="Chemical">alkynoates can
be realized using phosphine catalysis.[38] A recent example by Santos and co-workers involved PBu3 catalyzing the trans-phosphinoboration of internal alkynes 2–66 (Scheme ).[37b] The reaction
afforded trans-α-phosphino-β-boryl acrylates 2–67 in moderate to good yields with high
regio- and Z-selectivity, providing access to functionalized
phosphines that would be challenging to prepare otherwise.
Scheme 15
Trans-Phosphinoboration
of Internal Alkynes
In the presence of
nucleophile–electrophiles, class="Chemical">alkynes undergo
annulations to form a class="Chemical">pan class="Disease">diverse array of carbo- and heterocycles.[39] For example, the Kwon group reported the novel
phosphine-catalyzed annulation of 2-sulfonamidobenzaldehyes 2–68 and ynones 2–69 (Scheme ).[39b] The reaction was initiated through
α-umpolung addition of the alkyne 2–68 to form the ylide 2–71. Ensuing proton transfer, followed by an aldol reaction, generated
the annulation product benzo[b]azepin-3-one 2–70.
Scheme 16
Synthesis of Benzo[b]azapin-3-ones
Nucleophilic
Phosphine Catalysis of MBHADs
Together with class="Chemical">alkenes, class="Chemical">pan class="Chemical">allenes,
and alkynes, the MBHADs are also
excellent substrates for nucleophilic phosphine catalysis. The MBHADs 2–73 are typically activated by a tertiary
phosphine through SN2′ addition to give an active
phosphonium intermediate 2–74, which
undergoes various reactions when combined with nucleophiles, electrophiles,
and nucleophile–electrophiles (Scheme ). When a nucleophile is present, phosphine
catalysis of the MBHAD will involve a unique SN2′–SN2′ process leading to apparent SN2 allylation.
As is the case with allenoates, MBHADs can provide three-carbon units
in annulation reactions when treated with electrophiles. The known
MBHAD–alkene or MBHAD–azo [3 + 2] annulations continue
to be reported for asymmetric versions with more elaborate substrates.[40] MBHADs undergo the [4 + 1] annulation when mixed
with enone/enimine/diene electrophiles in the presence of a phosphine
catalyst. Several recent reports have demonstrated the incorporation
of various electrophiles, including 2-enoylpyridine (oxide)s, ortho-quinone methides, and α,β-unsaturated
imines.[41]
Scheme 17
Allylic Substitution
and Annulations with MBHADs
Alongside these developments, several new reactions have been class="Disease">disclass="Chemical">pan class="Chemical">covered
using more elaborate substrates.[42] In one
such example, the Huang group reported the domino [3 + 3] cyclization
of MBH carbonates 2–75 with nucleophile–electrophile para-quinamines 2–76 to
access hydroquinolines 2–77 (Scheme ).[42a] The reaction was initiated by the SN2′–SN2′ process to afford an allylic
amination intermediate 2–78. Subsequent
addition with the phosphine, followed by proton transfer, formed the
ylide 2–80. Intramolecular Michael
addition and elimination of the phosphine generated the hydroquinoline
product 2–77. In another example,
the same group disclosed a novel phosphine-catalyzed annulation of
an MBHAD with perfluoroalkylimidoyl chloride as the electrophile to
construct the perfluoroalkylated benzazepines 2–83 (Scheme ).[42b] Here, the phosphonium species generated
from the MBH carbonate and the phosphine was deprotonated by base;
it then added to the fluorinated imidoyl chloride to give the intermediate 2–84. Cyclization of the deprotonated 2–84 and subsequent enamine-to-imine tautomerization
afforded the perfluoroalkyl-substituted benzazepines. Notably, the
perfluoroalkylimidoyl chlorides provided four atoms of the benzazepine
products.
Scheme 18
Synthesis of Hydroquinolines
Scheme 19
Synthesis of Perfluoroalkylated Benzazepines
class="Chemical">Conventionally, class="Chemical">pan class="Chemical">MBH alcohols have been converted to their derivatives,
including acetates, carbonates, and halides. A recent report by Guo
and co-workers revealed that phosphine catalysis enables the direct
reaction of MBH alcohols with azomethine imines to access bicyclic
(epoxymethano)-pyrazolo[5,1-b]-quinazolines (Scheme ).[42c] They proposed that the MBH alcohol 2–88 was activated through reaction with the phosphine
to form the zwitterionic intermediate 2–90. The alkoxide of 2–90 underwent
sequential nucleophilic addition to the electron-deficient azomethine
imine 2–87, intramolecular Mannich-type
reaction, and SN2displacement of the catalyst to form
the bridged tetracycle 2–89.
Scheme 20
Phosphine-Catalyzed Reactions of Unmodified MBH Alcohols
Miscellaneous Nucleophilic
Phosphine Catalysis
In adclass="Disease">dition to activated C—C multiclass="Chemical">ple
bonds, tertiary class="Chemical">pan class="Chemical">phosphine
catalysts can also add to C=O and N=N groups and strained-ring
electrophiles, resulting in a diverse array of reactions. Several
recent examples are noteworthy. For example, the Behrouz and Rad group
reported an ultrasound-promoted PPh3-catalyzed three-component
condensation of benzil, urea, and aldehyde for the synthesis of 2,4,5-trisubstitutedimidazoles.[43] Recent progress in phosphine
catalysis has unveiled several unique rearrangements of strained cyclopropanes.[44] The Xu group demonstrated that phosphine-catalyzed
rearrangement of vinylcyclopropylketones can provide access to cycloheptenones.[44a] Subsequently, the Li and Xu group reported
the synthesis of tri- and tetrasubstituted furans, and trisubstituteddienones, through phosphine-catalyzed rearrangements of alkylidenecyclopropylketones
(ACPs, Scheme ).[44b] The development of this reaction was inspired
by a known palladium-catalyzed rearrangement of ACPs.[45] In this new study, three types of substrate-controlled
rearrangement of alkylidenecyclopropanes occurred to afford three
different product types. The results could be explained by considering
the allylic phosphonium zwitterions intermediates 2–101, 2–102, and 2–103 that formed upon nucleophilic addition of
the phosphine to the ACPs. When R1 was an alkyl group,
and R2 was an electron-withdrawing group, an SN2 process gave trisubstitutedfuran 2–96; when both R1 and R2 were aryl groups, an
SN2′ pathway formed tetrasubstituted furan 2–98; when R1 was a primary
alkyl group (RCH2) and R2 was an aryl substituent,
the intermediate 2–103 underwent
elimination and the alkene isomerization to afford the trisubstituteddienone 2–100.
Scheme 21
Phosphine-Catalyzed
Rearrangements of Alkylidenecyclopropylketones
Summary
Nucleophilicclass="Chemical">phosphine catalysis
has remained one of the most vigorous areas of research in the field
of class="Chemical">pan class="Chemical">phosphorus-based organocatalysis. It continues to expand its horizons
to the assembly of carbo- and heterocycles through new modes of reaction.
Recent progress in this area has shared some common features. Unforeseen
reactivity patterns continue to be discovered when highly functionalized
substrates, deviating from generic versions, are used in classical
reactions (e.g., Schemes and 6). Enantioselective variants
of the classical nucleophilic phosphine catalysis processes continue
to be reported (e.g., Schemes and 8). Brand new reactions continue
to be uncovered, often accompanied by asymmetric variants when using
chiral phosphines (e.g., Schemes and 10). The application of
nucleophilic phosphine catalysis is a promising approach for the syntheses
of functional molecules (e.g., Schemes and 7). In addition, the zwitterionic
adducts generated in situ from nucleophilic addition of a phosphine
to electron-deficient multiple bonds can be efficient basic catalysts
for asymmetric Mannich-type reactions, Strecker reactions, Michael
additions, and aldol reactions.[46] In a
recent example, Wang and co-workers revealed that the zwitterion 2–109 generated from the multifunctional
chiral phosphine 2–107 and the allenoate 2–108 was highly efficient in the asymmetric
decarboxylative Mannich reactions between the cyclic ketimines 2–104 and the β-keto acids 2–105 (Scheme ).[47] Notably,
the reactions could be performed to give high yields and excellent
enantioselectivities (from 99.0% to >99.9% ee) at a very low catalyst
loading (0.5 mol %). Considering the ready availability of chiral
phosphines and electron-deficient alkenes, allenes, and alkynes, this
strategy may open a new door for phosphonium-salt catalysis in asymmetric
synthesis. Lastly, with its diverse reaction modes, we envision that
nucleophilic phosphine catalysis will have many more applications
in the syntheses of functional molecules—including biologically
active natural products, pharmaceuticals, agrochemicals, and materials.
Whereas the class="Chemical">phosphorus reagents toggle between the tertiary class="Chemical">pan class="Chemical">phosphine(III)
state and quaternary phosphonium(V) species in nucleophilic phosphine
catalysis, classical phosphorus(III)-based processes (e.g., Wittig,
Mitsunobu, Appel, and Cadogan cyclization reactions) are driven by
the formation of strong P(V)=O double bonds. The formation
of a phosphine oxide as waste is, however, an issue that has troubled
chemists for a long time. Although the issue can be partly solved
through immobilization and precipitation of the phosphine oxide,[48] catalytic cycling of the phosphorus species
would be more atom-economical and environmentally friendly.[49] Two strategies, namely, redox-neutral and redox-driven
phosphine oxide catalysis, are commonly adopted to recycle phosphorus
species in such reactions (Scheme ).[49,50] The redox-neutral catalysis requires
in situ activation of the phosphine oxide to form an active P(V) intermediate,
which will further react with substrates and regenerate phosphine(V)
oxide. In the redox-driven catalysis, P(III) usually reacts with substrate 3–1 to form an active P(V) intermediate,
which will further react with the substrate 3–2 to give the product and the phosphine oxide. To turn over
the phosphine oxide to P(III), stoichiometric reductant is needed.
The harsh conditions required for the activation or reduction of the
inert P(V)=O double bond, however, have largely plagued the
development of catalytic processes. In the past decade, significant
advances have been made in this field.[49] Examples include the development of catalytic Wittig-type reactions
using arsine or telluride ylides,[51] gas
evolution (N2, CO2) to assist the redox-neutral
catalytic process,[49,50,52] and the use of silanes as reductants in the redox-driven catalytic
cycle.[49,53] Several new advances have emerged since
our review in 2018,[3b] and we highlight
them below.
Scheme 23
Redox-Neutral and Redox-Driven Phosphine Oxide Catalysis
Redox-Neutral Phosphine Oxide Catalysis
In redox-neutral class="Chemical">phosphine oxide catalysis, the class="Chemical">pan class="Chemical">phosphine oxide
is converted directly into another reactive P(V) species. For example,
treatment of a phosphine oxide with an isocyanate or oxalyl chloride
produces iminophosphorane[52] or chlorophosphonium
chloride[50] intermediates that have been
used in the aza-Wittig or Appel reaction and in the dehydration of
oximes to nitriles. The use of this mode of phosphine oxide catalysis,
although known since 1962, has lagged behind the thriving activities
of redox-derived catalysis.[52a] Recently,
however, in what could be considered one of the most exciting innovations
of late, Denton and co-workers developed a redox-neutral organocatalytic
Mitsunobu reaction (Scheme ).[54] The Mitsunobu reaction is
commonly used for the nucleophilic substitution of primary and secondary
alcohols with inversion of their chiral centers in a stereospecific
manner. Despite its usefulness, the requirement of stoichiometric
amounts of reagents (PPh3, azo oxidant) and the generation
of undesired waste have undermined this powerful transformation. Although
several attempts have been made to develop a redox-driven organocatalyzed
Mitsunobu reaction, a stoichiometric amount of reductant has been
required to recycle the phosphine oxide, and a stoichiometric amount
of oxidant has been required to regenerate the azo oxidant in cases
where a substoichiometric amount of the azo species is used.[49f] In Denton’s study, a phosphine oxide
catalyst 3–5 was employed without
any other additives. Notably, the previously required azo oxidant
is not necessary here, because the active dihydrobenzooxaphospholium
intermediate 3–6 forms with the evolution
of benign water as the sole byproduct. Subsequent substitution by
the alcohol 3–3 generates the alkoxyphosphonium
intermediate 3–7, which then undergoes
another substitution with the nucleophile anion to afford the product 3–4, while simultaneously regenerating
the catalyst 3–5. By maintaining
a constant oxidation state of the phosphorus species, this approach
eliminates any need for stoichiometric oxidants/reductants and produces
water as the sole byproduct. This process is both scalable and compatible
with a variety of pronucleophiles, forming C–O, C–N,
and C–S bonds. Furthermore, protocols have been developed to
enable one-pot activation of simple alcohols (en route to the synthesis
of API and symmetrical ethers) when using p-toluenesulfonic
acid or trifluoromethanesulfonic acid as a cocatalyst.
Scheme 24
Catalytic
Mitsunobu Reaction
Redox-Driven
Phosphine Oxide Catalysis
Undoubtedly, O’Brien’s
2009 report on a catalyticclass="Chemical">Wittig
reaction, in which a class="Chemical">pan class="Chemical">silane reduces the phosphine oxide byproduct
back to phosphine in situ during the reaction, unleashed the imaginations
of researchers around the world.[53] Consequently,
the subsequent decade witnessed numerous accounts of catalytic versions
of classical P(V)=O-generating reactions, including Wittigolefination, Appel reaction, Staudinger reduction, Staudinger ligation,
aza-Wittig reaction, Mitsunobu reaction, and Cadogan cyclization,
with all employing stoichiometric silanes as terminal reductants.[3b,49] While studies of redox-driven phosphine oxide catalysis of the aforementioned
reactions have continued,[55−59] the past two years have been marked by the application of redox-driven
phosphine oxide catalysis to lesser-known reactions.
One of
the reactions that is meclass="Disease">diated by a class="Chemical">pan class="Chemical">phosphine, but not widely known,
is the cross-coupling between nitrosoarenes and boronic acids for
the synthesis of di(hetero)arylamines, reported by Csákÿ.[60] In 2018, the Radosevich group disclosed a deoxygenative
C–N crosscoupling between nitroarenes and boronic acids (Scheme ).[61a,61b] While transition-metal-mediated C–N couplings are used widely,
the P(V)/P(III)-catalyzed reaction is a complementary main group redox-catalysis
method. In particular, this transformation tolerates various functional
groups, including a series of heterocycles and halogen substituents.
Radosevich et al. proposed a two-stage deoxygenation sequence for
the reductive C–N coupling. Initially, [3 + 1] cheletropic
addition of the nitrobenzene 3–8 and
the phosphetane 3–12 generated the
intermediate 3–13, decomposition
of which furnished nitrosoarene and the oxide 3–11. The second deoxygenation began with addition of the phosphetane
to the nitrosoarene to form the betaine intermediate 3–16, which underwent 1,2-migration to give the
desired product with concomitant generation of the phosphetaneoxide 3–11. The Radosevich group recently demonstrated
that nitromethane is amenable to the reaction, realizing methylamination
of arylboronic acids.[61c] When the nitroarenes
have different ortho substituents, sequential reductive coupling and
cyclization could occur efficiently to yield various heterocycles
(Scheme ).[62] A tandem coupling/acylation would generate oxindoles 3–18 and quinoxalinediones 3–19, while a tandem coupling/condensation would
form indoles 3–20 and benzimidazoles 3–21.
Scheme 25
Reductive C–N Cross-Coupling between
Nitroarenes and Boronic
Acids
Scheme 26
Synthesis of Azaheterocycles through
Reductive Coupling between Nitroarenes
and Boronic Acids
class="Chemical">Comclass="Chemical">pared with the
oxidation of class="Chemical">pan class="Gene">S(II) species into valuable sulfoxides
and sulfones, the deoxygenative transformation of sulfur-containing
compounds is less recognized as synthetically useful. The Sharpless
group had previously demonstrated that trialkylphosphitecould reduce
sulfonyl chlorides to corresponding sulfinyl chlorides and, subsequently,
to sulfenyl chlorides, and that the sulfinyl chloridecould be captured
by nucleophiles (e.g., alcohols) to form sulfinate esters.[63a] The final step in their proposed mechanism
for the sequential deoxygenation of sulfonyl chloride was the reaction
of the sulfenyl chloride with trialkylphosphite to form trialkoxyl(arylthiol)phosphonium
chloride, which underwent a Michaelis–Arbzov reaction to produce
the corresponding phosphorothiolate and alkyl chloride.[63b] Recently, Radosevich and co-workers reported
that the phosphetaneoxide 3–11 was
capable of catalyzing electrophilic sulfenylations through deoxygenation
of sulfonyl chlorides with indoles as the nucleophile (Scheme ).[64] The reaction is suitable for a diverse range of indoles and sulfonyl
chlorides, providing access to indole derivatives. For this 2-fold
deoxygenative sulfenylation, 3–25 was proposed as the key intermediate responsible for the electrophilic
substitution. Again, the P(V)/P(III) cycle was driven by phenylsilane-mediated
reduction of P(V) oxide to P(III).
Scheme 27
Electrophilic Sulfenylation
through Deoxygenation of Sulfonyl Chlorides
class="Chemical">Halophosphonium cations have class="Chemical">proven to be effective for deoxygenation
in both the Aclass="Chemical">pclass="Chemical">pel reaction and class="Chemical">pan class="Chemical">amide formation.[65] This behavior inspired the development of phosphetaneoxide 3–11-catalyzed didehydration for the preparation
of functionalized heterocyclic adducts (Scheme ).[66] The annulation
was amenable to both aniline and benzylamine for the construction
of six- and seven-membered rings. The key to the success of this sequential
annulation is the generation of the halophosphonium cation 3–29 from the in situ-generated phosphetane and
the oxidant diethyl bromomalonate (DEBM). Activation of a carboxylic
acid with the halophosphonium cation, followed by coupling, resulted
in the first dehydration to form an amide intermediate. The amide
bond was further activated by the halophosphonium species, with the
ensuing SEAr cyclodehydration generating the desired heterocycles.
Scheme 28
Annulation of Amines and Carboxylic Acids through Iterative Dehydration
class="Chemical">Alkene reduction is one of the most fundamental
reactions in organic
synthesis. While transition class="Chemical">pan class="Chemical">metal catalysis dominates the field, organocatalysis
approaches may complement the former by providing unique selectivity
and functional group compatibility. Indeed, several successful examples
of phosphine-mediated reduction of activated double and triple bonds
have been reported.[67] These reductions
have involved addition of the phosphine to the double or triple bond
to form the zwitterion 3–32 and a
subsequent reaction with water to form the phosphorane 3–33, which collapsed to the product 3–31 and the phosphine oxide (Scheme ). A stoichiometric quantity
of the phosphine was, however, required, producing undesired phosphineoxide waste. Last year, the Werner group reported the first example
of a phosphine oxide-catalyzed alkene reduction through silane-mediated
P(III)/P(V)=O redox cycling.[68] In
this transformation, a combination of the strained phosphetaneoxide 3–11 and phenylsilane was used to replace
the previously employed stoichiometric phosphine. The reduction selectively
targeted the activated double bonds without affecting other functional
groups susceptible to hydrogenation. Remarkably, water was tolerated
by the silane-promoted P(III)/P(V)=O redox cycling.
Scheme 29
Reduction
of Activated Alkenes through P(III)/P(V)=O Redox
Cycling Catalysis
Asymmetric
Phosphine Oxide Catalysis
Although chiral class="Chemical">phosphines have
been reclass="Chemical">pan class="Chemical">cognized as reagents for asymmetric
Wittigolefination and Staudinger–aza-Wittig reactions, explorations
in this field have been limited by the noneconomical nature of those
reactions.[69] Today, significant advances
in phosphine oxide catalysis have given an impetus for renewed interest
in this area, such that enantioselective catalysis processes driven
by the P(III)/P(V)=O redox cycling of chiral phosphine oxides
are emerging as powerful tools for the synthesis of chiral molecules.
In particular, desymmetrization of meso compounds is a valuable strategy
for accessing molecules with high optical purity. In 2014, Werner
et al. reported the first enantioselective catalytic Wittig reaction
through desymmetrization of 2-(3-bromo-2-oxopropyl)-2-methylcyclopentane-1,3-dione
with catalytic Me-DuPhos (3–37) and
phenylsilane as the terminal reductant.[70a] They found that Me-DuPhos worked efficiently for asymmetric induction
with up to 90% ee, but with less than 10% yield. Three years later,
the Christmann group reinvestigated this reaction and found that the
desymmetrizing Wittig reaction could provide high enantiomeric excess
(92–96% ee) and moderate yields (Scheme ).[70b] They found
that premature debromination of the triketone, caused by impurities
in the starting materials, might have been responsible for the low
yield reported initially. Although high temperature (150 °C)
was required for the catalytic intramolecular Wittig reaction, good
enantioselectivities were obtained when using Me-DuPhos. The optically
pure bicyclic enone products were further elaborated into various
carbocyclic propellanes through sequential Michael addition with Lipshutz’s
cyanocuprates and ring-closing metathesis.[70b] The propellane 3–39 was applied
to the first asymmetric syntheses of ent-dichrocephone
A and ent-dichrocephone B.[70c]
Scheme 30
Synthesis of ent-Dichrocephone A and ent-Dichrocephone B
Although good selectivity was obtained in some cases for the asymmetricclass="Chemical">Wittigclass="Chemical">pan class="Chemical">olefination at high temperature, mild room-temperature reactions
would be more desirable. The inert nature of P(V)=O poses a
challenge for its silane-mediated reduction to a trivalent phosphine
at ambient temperature. Nevertheless, continued studies in this field
have unveiled some factors for lowering the energy barrier of the
silane-mediated phosphine oxide reduction, such as the use of the
strained small-ring phosphetaneoxide 3–11 as the catalyst. Interestingly, strained phosphine oxides
can be reduced more readily than unstrained ones. A recent computational
study explained why strained phosphetanes and bridged [2.2.1] bicyclicphosphines are privileged structures in redox-driven phosphine oxide
catalysis.[71] One reason is that one of
the C—P—C angles of these phosphines is close to that
in the proposed transition state for the intramolecular hydride delivery
in the (phosphoniooxy)silcate(IV) adduct formed between the phosphineoxide and the silane. Studies have also demonstrated that both acidic
and basic additives can facilitate the silane-mediated phosphine oxide
reduction.[72] Consolidating these lessons,
in 2019, Kwon and co-workers developed a catalytic asymmetric Staudinger–aza-Wittig
reaction for the synthesis of chiral cyclic imines 3–44 (Scheme ).[73] In this reaction, the commercially
available P-chiral [2.2.1] bicyclicendo-phenyl-HypPhos 3–45 was used to
facilitate desymmetrization of the two C=O groups of various
1,3-diketones and to form enantioenrichediminoketonescontaining
a chiral quaternary center. The strained nature of the catalyst, in
conjunction with the 2-nitrobenzoic acid additive, enabled room-temperature
redox cycling mediated by phenylsilane.
In another recent report, the Voituriez group documented an efficient
method to access class="Chemical">enantioenrichedclass="Chemical">pan class="Chemical">(trifluoromethyl)cyclobutenes 3–48 through an asymmetric α-umpolung
addition–Wittigolefination cascade between the dione 3–46 and the acetylenedicarboxylate 3–47 (Scheme ).[74] The catalyst
employed here, exo-anisyl-HypPhos oxide (3–49), is also a strained [2.2.1] bicyclicphosphineoxide, which, along with the phosphoric acid additive, facilitated
mild P(III)/P(V)=O redox cycling. This tandem reaction represents
another powerful example of phosphine oxide redox cycling as a tool
for the synthesis of attractive chiral molecules.
Scheme 32
Enantioselective
Synthesis of (Trifluoromethyl)cyclobutenes
Although significant progress
has been made, challenges remain in the area of class="Chemical">phosphine oxide catalysis.
Most redox-driven class="Chemical">pan class="Chemical">phosphine oxide catalysis has been based on silane-mediated
in situ reduction. The need for a stoichiometric amount of reductant
raises economic and environmental concerns.[75] A recent study by Sevov and co-workers revealed that the triphenylphosphineoxide byproduct of the Wittigolefination can be transformed directly
into PPh3 through electroreduction (Scheme ).[76] Revitalized
adaptation of such unconventional technologies may open new doors
for redox-driven phosphine oxide catalysis. Conversely, as demonstrated
by Denton and co-workers in their redox-neutral organocatalytic Mitsunobu
reaction, designing novel catalysts may also be a promising avenue
for green chemistry. Although catalytic versions of those phosphineoxide-driven reactions are appealing, unless their selectivities and
yields are comparable with those of the original versions, their utility
will be limited. In this regard, there is a great need to develop
mild methods for reduction or activation of phosphine oxides. Accordingly,
asymmetric phosphine oxide catalysis is another appealing field benefiting
from ongoing progress in phosphine oxide catalysis. Although much
needs to be done, phosphine oxide catalysis finds itself at the threshold
of a bright future.
Scheme 33
One-Pot Wittig Olefination/Electroreduction
Phosphorus-Mediated Radical
Processes
class="Chemical">Organophosphorusclass="Chemical">pan class="Chemical">compound-mediated SET processes
have received
much less attention than their nitrogencounterparts,[77] despite trivalent phosphine radical cations and phosphoranyl
radicals having been recognized as reaction intermediates as early
as the 1950s.[78] Progress in this emerging
field has established that organophosphoruscompounds can act as radical
initiators, catalysts, and radical precursors in various SET transformations
for the synthesis of important structural motifs.[9,79] With
the rapid development of photocatalysis in recent years, the use of
organophosphoruscompounds as radical precursors, especially for the
purpose of deoxygenation or desulfurization, is becoming established
as a reliable method.[9] Several examples
of organophosphoruscompound-mediated radical processes are discussed
below.
Use of Substoichiometric Phosphine
Electron
donor–acceptor (class="Chemical">EDA) class="Chemical">pan class="Chemical">complexes, also known as charge-transfer
complexes (CTCs), are typically formed from electron-rich and electron-poor
compounds and possess distinct properties differing from those of
their parent compounds, enabling some unique transformations.[80] In particular, even when the parent compounds
cannot absorb visible light, their EDAcomplexes can, facilitating
useful transformations.[81] In step with
recent attention paid to the study of EDAcomplexes, especially in
the area of photoredox chemistry, the use of phosphoruscompounds,
which are good electron donors for forming EDAcomplexes, is gathering
momentum. In 1990, the Huang group discovered that a catalytic amount
of PPh3could promote the perfluoroalkylation of simple
alkenes with perfluoroalkyl iodides through a radical chain reaction
(Scheme ).[82] In this transformation, an EADcomplex 4–4, formed from PPh3 and perfluoroiodine 4–1, would decompose to the radical intermediate
Rf•. Addition of Rf• to an alkene
generates the alkyl radical intermediate 4–5. The reaction of 4–5 with
perfluoroiodine forms the product 4–3 and regenerates Rf• for radical chain propagation.
While subsequent examinations of this reaction were conducted under
more or less the same thermal conditions as those in Huang’s
original study,[83a−83c] Czekelius and co-workers found in 2019 that
light also facilitated activation of the EADcomplex, allowing the
reaction to occur at close to ambient temperature (Scheme ).[83c] A systematic study revealed that Bu3P and light from a blue LED (461 nm) were optimal for promoting
this radical chain reaction.
Scheme 34
Phosphine-Initiated Perfluoroalkylation
of Alkenes
In the class="Chemical">EDAclass="Chemical">pan class="Chemical">complex-initiated
radical processes discussed above,
the substrates are involved in the EDAcomplexes and incorporated
into the final products. This strategy requires careful selection
of substrates to match the electronic properties of the phosphines,
limiting their scope and, consequently, the utility of this strategy.
A more general and effective strategy, however, may involve a CTC
catalyst that can transfer an electron to a substrate to assist the
radical generation and later accept an electron from the late-stage
intermediate to complete the catalytic cycle. This strategy was realized
by the Fu group in 2019.[84] Although most
photoredox catalysts involve precious metalcomplexes or organic dyes,
they demonstrated a new type of photocatalytic decarboxylative alkylation
using the combination of PPh3 and sodium iodide (Scheme ). This novel alkylation
was realized with several types of alkyl radical precursors (decarboxylative
alkylation via redox-active esters, deaminative alkylation via Katritky’s N-alkylpyridinium salts, and trifluoromethylation with Togni’s
reagent) and radical acceptors (e.g., silyl enol ethers, activated
heterocycles, and 1,1-disubstituted alkenes). In the proposed mechanism,
Fu et al. suggested that the CTC 4–10 is formed initially through Coulombic interactions between PPh3, sodium iodide, and the N-(alkylcarbonyl)phthalimide 4–8. SET from iodide to the phthalimide
moiety then prompts decomposition and concomitant generation of an
alkyl radical, which adds to the radical acceptor. Finally, the transient
radical cation complex 4–12 oxidizes
the amine radical cation 4–15 to
the product 4–16 and turns back to
the PPh3–NaIcomplex 4–10. A chiral phosphoric acidcould enable enantioselective
Minisci-type reactions with high levels of stereoselectivity. Interestingly,
the blue LED light (456 nm) used here is very similar to that (461
nm) used in the perfluoroalkylation of alkenes (Scheme ), suggesting that related
complexes might have formed.
Scheme 35
Photocatalytic Decarboxylative Alkylations
Mediated by PPh3 and NaI
In a related study, the He, class="Chemical">Xue, and Chen grouclass="Chemical">ps reclass="Chemical">ported a class="Chemical">pan class="Gene">radical
C–H arylation of oxazoles with aryl iodides.[85] They proposed that the complex D formed from
Cs2CO3 and 1,1′-bis(diphenylphosphino)ferrocene
(dppf) could serve as a strong electron donor, reducing aryliodine
to form the radical anion intermediate 4–20 and further initiating a radical chain reaction (Scheme ). The radical 4–20 would undergo C–I bond cleavage
to form the aryl radical 4–21. Addition
of 4–21 to oxazoles, followed by
deprotonation, would form the radical anion 4–24. Reduction of the aryl iodide by the radical anion 4–24 would form the product 4–25 and regenerate the radical anion 4–20 for radical chain propagation. The radical
anion 4–24 could also reduce D(+1) to D and complete the redox cycle. DFT
calculations indicated, however, that dissociating CsCO3– from D(+1) to generate a potential
electron acceptor (dppf)+ is thermodynamically challenging.
The authors favored the radical chain reaction pathway, while not
ruling out the redox cycle pathway.
Scheme 36
dppf/Cs2CO3-Mediated Oxazole C–H Arylation
In adclass="Disease">dition to class="Chemical">pan class="Chemical">complexing with other compounds to form
catalysts
for SET processes, phosphoruscompounds themselves can also act as
electron shuttles in redox cycles. In 2015, Tan and Liu employed Togni
reagent II as the source of trifluoromethyl radicals for the phosphine-catalyzed
bistrifluoromethylation of enamides (Scheme ).[86a] The reaction
involved multiple SET processes, with mechanistic studies indicating
that the catalytic cycle was driven by redox cycling of trivalent
phosphine and a phosphorus-centered radical cation. Initially, the
phosphine 4–28 was oxidized by the
Togni reagent II through SET to form the CF3 radical and
the phosphorus-centered radical cation 4–30. Addition of the CF3 radical to the alkene,
followed by a 1,5-hydride shift, generated the α-amidobenzyl
radical 4–33. Subsequent oxidation
by the Togni reagent II and deprotonation formed the enamide 4–35. Addition of the CF3 radical
to 4–35 afforded the radical intermediate 4–36, which was further oxidized by the
phosphorus radical cation and deprotonated to give the product 4–38. The SET oxidation also recycled
the phosphine catalyst 4–28. This
method was further developed by the authors to allow concomitant trifluoromethylation
of the alkene and subsequent remote alcohol or amine α-C–H
activation to afford valuable trifluoromethyl-substituted ketones
and aldehydes.[86b]
In
adclass="Disease">dition to their roles as class="Chemical">pan class="Gene">radical initiators and catalysts as described
above, trivalent phosphorus(III)compounds are increasingly being
used as radical precursors for SET transformations.[9] Although, at present, stoichiometric reagents (e.g., as
ylide or phosphine) are required, the byproducts formed in the deoxygenation
or desulfurization process may be recycled in situ, as we have discussed
above regarding phosphine oxide catalysis. Mechanistically, fragmentation
of phosphoranyl radicals, through either α- or β-scission,
is an efficient means of forming radicals for further transformation.
The mechanisms of formation of phosphoranyl radicals can be distinguished
by whether they involve addition of an external radical to a phosphine
or addition of a nucleophile to a phosphorus-centered radical cation,
and also by whether light is involved in the formation of the phosphoranyl
radicals. Because this emerging field has been summarized in two recent
reviews,[9] we discuss only selected examples
in this Outlook, including the α- and β-scissions of phosphoranyl
radicals, light-associated and non-light-associated SET processes,
as well as two modes of phosphoranyl radical generation.
Miura
and class="Chemical">co-workers reclass="Chemical">ported recently the class="Chemical">photoredox-driven fragmentation
of class="Chemical">pan class="Chemical">phosphoranyl radicals through α-scission to generate active
radical species.[87] They disclosed a practical
and efficient synergistic photoredox approach for the preparation
of elongated esters from alkenes and stabilized phosphorus ylides
under mild conditions (Scheme ). The phosphoranyl radical intermediate generated
through [IrIII*]-mediated reduction of the phosphonium
salt 4–42 further fragmented into
PPh3 and the radical 4–44 for subsequent addition to alkenes. Notably, the phosphoranyl radical
was generated through neither of the methods described above.
Scheme 38
Synthesis of Elongated Alkenes
The Zhu and Xie group provided an example of photoredox-driven
fragmentation of class="Chemical">phosphoranyl radicals through β-scission in
their efficient aclass="Chemical">pclass="Chemical">proach to class="Chemical">pan class="Chemical">ketones through the deoxygenation of aromatic
carboxylic acids using synergistic photoredox catalysis in the presence
of PPh3 (Scheme ).[88a] The reaction was initiated
through oxidation of PPh3 to its radical cation, mediated
by a photoexcited catalyst. The radical cation coupled with the carboxylate
anion to generate the phosphoranyl radical intermediate 4–46, the β-scission of which generated
the acyl radical 4–47. The final
product was formed through radical addition to an alkene. In addition,
corresponding deuterated aldehydescould be formed by the acyl radical
intermediate abstracting a deuterium atom from D2O.[88b] Concurrently, the Doyle group reported the
deoxygenation of alcohols and carboxylic acids 4–48 through a similar phosphine-promoted photoredox-catalysis
reaction.[88c]
Scheme 39
Photoredox-Driven
Fragmentation of Phosphoranyl Radicals through
β-Scission
Whereas the examples
above involving class="Chemical">phosphorusclass="Chemical">pan class="Chemical">compound-mediated
SET were driven by photocatalysis, phosphoranyl radical intermediates
can also be formed under thermal conditions. In 2018, the Schmidt
group reported an intermolecular anti-Markovnikov hydroamination of
alkenes using triethylphosphite and N-hydroxyphthalimide
(Scheme ).[89] The reaction was initiated by hydrogen atom
abstraction from N-hydroxyphthalimide to produce
PhthNO•, which added to triethylphosphite to form the phosphoranyl
radical intermediate 4–54. Subsequent
β-fragmentation generated the radical PhthN•, which added
to the alkene 4–51 to generate the
radical 4–56. Abstraction of a hydrogen
atom from 4–50 formed the anti-Markovnikov
alkene hydroamination product 4–52 and regenerated PhthNO•.
Scheme 40
Anti-Markovnikov Alkene Hydroamination
class="Chemical">Organophosphorusclass="Chemical">pan class="Chemical">compound-promoted
SET reactions have reactivity complementary to that of nucleophilic
phosphine catalysis. In nucleophilic phosphine catalysis, the substrates
are limited to electron-deficient alkenes, whereas inert alkenes (as
demonstrated in Schemes and 37–40) and C–H bonds (as demonstrated in Schemes –37) can be
activated in SET processes. Special attention may be given to the
EDAcomplexes of phosphoruscompounds, which are sensitive to light
and can either initiate or catalyze SET processes. Exploration of
the EDAcomplexes of phosphoruscompound should unveil rich chemistry
in this underdeveloped area. A special type of EDAcomplex, the FLP,
can be a powerful tool in the activation of small molecules. Although
a two-electron-transfer mechanism is generally recognized,[6b] recent reports have indicated that SET processes
might also play important roles in FLP chemistry.[90,91] An interesting study by Stephan and co-workers unveiled that a frustrated
radical pair (FRP) [Mes3P•]/[•E(C6F5)3] (E: B/Al) could be formed through a SET
process, while the analogous tBu3P pair
behaved differently without undergoing such a process.[91a] This finding, as well as others, provides evidence
that the structural features of tertiary phosphines greatly influence
the stability of phosphorus-centered radical cations and can even
alter the reaction pathway.[91] Moreover,
a tertiary phosphine itself can function as an electron shuttle in
a redox-catalytic cycle. Organophosphoruscompound-promoted SET processes,
therefore, have huge potential for applications in photochemistry
and electrochemistry. The limited number of reports of organophosphoruscompound-catalyzed SET reactions may be due to the short half-lives
of trivalent phosphine radical cation species.[91d] Designing phosphines or their complexes with structural
frameworks of suitable electronic potential may result in longer half-lives
for the open-shell species and benefit this area. The unique role
of organophosphoruscompounds to generate radical intermediates through
SET processes is also appealing.
Miscellaneous
Catalyses
In adclass="Disease">dition to nucleoclass="Chemical">philic class="Chemical">pan class="Chemical">phosphine catalysis,
catalysis to bypass
phosphine oxide waste, and phosphorus-mediated radical processes,
organophosphoruscompounds are used in multitudes of catalytic processes,
including phosphonium salt phase transfer catalysis (PTC), iminophosphoranesuper base catalysis, asymmetric phosphine oxide Lewis base catalysis,
FLP catalysis, and chiral phosphoric acid catalysis. PTC is considered
a powerful and green sustainable tool for addressing synthetic challenges.[7] In this context, phosphonium-containing organocatalysts,
especially bifunctional catalysts, are emerging as powerful phase
transfer catalysts for various asymmetric transformations (e.g., alkylation,
Michael addition, Darzens reaction, Mannich reaction, and Strecker
reaction).[7] Deprotonation of pronucleophiles
with chiral Brønsted bases (e.g., cinchona alkaloids and derivatives)
to induce enantioselective reactions via ion pairs is an established
reliable method for realizing various enantiodifferentiating transformations.
Although powerful, the application of this method to high-pKa pronucleophiles has been limited. The availability
of phosphorus-based super bases [e.g., P-spiro chiral triaminoiminophosphorane,
chiral bis(guanidino)iminophosphorane, and bifunctional iminophosphorane],
with values of pKBH+ ranging
from 22 to 32.9, has opened a new door in this area, realizing various
challenging reactions that are difficult to perform using cinchona-based
catalysts.[8] Another type of phosphorus-related
base catalysis is asymmetric phosphine oxide Lewis base catalysis.
Generally, a polarized phosphine oxide bond coordinates to chlorosilanes
to form pentavalent siliconcomplexes, thereby enhancing the electrophilicity
of the silicon species to activate electrophiles (e.g., carbonyls,
epoxides, and imines) for further transformation. In the presence
of chiral phosphine oxides (e.g., chiral binaphthyl dioxides and chiral
phosphoramides), a variety of asymmetric transformations (e.g., allylation,
aldol, hydrophosphonylation, and reduction of unsaturated compounds)
can be promoted.[5]
Alongside base
catalyses, class="Chemical">phosphorus-based Brønsted acid catalysis
has also beclass="Chemical">pan class="Chemical">come established as a reliable and efficient method to
form a variety of C–C, C–H, and C–X bonds enantioselectively.[4] Since the introduction of chiral phosphoric acid
into catalytic asymmetric reactions in 2004,[92] the following years have witnessed tremendous progress, resulting
in the development of a variety of chiral acids (e.g., BINOL derivatives,
TADDOL derivatives, VAPOL derivatives, and spiro phosphoric acids).[4] Today, phosphorus-based Brønsted acid catalysis
remains an active area of inquiry. Finally, FLPs, formed from combinations
of Lewis acids and Lewis bases, have been used in the development
of main group element catalysis processes for the activation of small
molecules (e.g., H2, CO2, CO, SO2).[6] In this context, various phosphines
have played important roles in combination with Lewis acids [e.g.,
B(C6F5)3 and Al(C6F5)3] to form active FLPs.[6] Although such studies constitute a large genre of the field of organophosphorus-based
catalysis, we cannot cover them in this Outlook, due to limited space;
therefore, we direct the reader’s attention to recent reviews
on each topic.[4−8]
Although the sclass="Chemical">coclass="Chemical">pe of class="Chemical">pan class="Chemical">organophosphorus-based catalysis is
large,
most of its topics are “young” in that they have been
established as fields of systematic inquiry only within the last 20
years. For example, redox-driven phosphine oxide catalysis began to
gain significant interest in 2009 after O’Brien published the
catalytic Wittig reaction;[53] the phosphoruscompound-mediated SET process has been attracting more attention recently
as a result of the rapid development of photoredox chemistry;[9] FLP chemistry became popular only after Stephan
reported, in 2006, that H2 can be activated by an FLP.[93] Notably, one of the original functions of chiral
phosphoric acid, first reported in 1971, was as a stoichiometric reagent
for the resolution of racemic amines.[94] The rich chemistry of phosphoruscompounds will enable more interesting
reactions to be added to the toolbox of organic synthesis. One lesson
we can take here is that new chemistry is emerging and that previous
stoichiometric reactions may turn into catalytic ones in due course.
Although not catalytic in its organophosphorus reagents, phosphorus(V)
ligand coupling has been gaining renewed interest recently. We discuss
these so-called “contractive coupling” reactions briefly
in the next section.
Phosphorus(V) Ligand Coupling
Like the other main group elements below the seclass="Chemical">cond row of the
class="Chemical">perioclass="Chemical">pan class="Disease">dic table (e.g., S, I, Si), phosphorus can adopt hypervalent
states featuring an expanded valence-shell, tending to extrude an
electron pair to form a more stable octet.[95] The collapse of the hypervalent species results in various transformations,
one of which is ligand coupling.[96] The
notion of “ligand coupling” was first introduced by
Oae to describe the exclusion of two ligands from hypervalent species
to form a stable octet and the coupling product.[96,97] It was also termed “contractive coupling” by Newkome
earlier in describing a similar process for the synthesis of biheteroaryls.[98] While the concept has been applied to various
main group elements,[97] recent progress
has demonstrated that P(V)-mediated ligand coupling can be a powerful
method for heterocycle functionalization.[99] Many heterocycles, including pyridines, diazines, and benzothiazoles,
are important pharmacophores. While transition-metal-mediated functionalization
of aromatic rings and electrophilic aromatic substitution are very
well established, functionalization of these heterocycles remains
a big challenge in terms of site-selectivity and functional group
compatibility.[100] In this context, phosphorus
ligand coupling has found itself very useful in medicinal chemistry.
Although a stoichiometric amount of reagent is required at present
for phosphorus ligand coupling, the contractive coupling process mimics
reductive elimination in transition-metal-catalyzed coupling (Scheme ). Recently, the
main group element-mediated two-electron redox catalysis has demonstrated
its potential in coupling reactions.[101] The Radosevich group reported a T-shaped P(III)compound as the
catalyst for transfer hydrogenation of azobenzenes,[101b] and Cornella and co-workers unveiled a novel Bicomplex
that facilitates the fluorination of arylboronic esters through redox
catalysis.[101c,101d] Such progress sheds light on
the potential for catalytic phosphorus-mediated ligand coupling. Below,
we discuss some recent advances in the area of phosphorus ligand coupling.
Scheme 41
Transition-Metal-Catalyzed Coupling and Phosphorus(V) Ligand Coupling
The phenomenon of class="Chemical">couclass="Chemical">pling between two ligands
has been observed
frequently at the class="Chemical">pan class="Chemical">phosphorus center.[97,98,102,103] For example, phosphineoxides or phosphonium salts bearing two or three 2-pyridyl groups
can undergo contractive ligand coupling to afford bipyridines.[98,102b,102c] The introduction of the heterocycles
at the phosphorus center was, however, the most challenging step in
those reports. In Newkome’s study, phenyldipyridylphosphine
(oxide)s were prepared through reactions between dilithiophenylphosphide
and halopyridines or through addition of 2-lithiopyridines to dichlorophenylphosphine.[98] The success of ligand coupling requires a reliable
method to access the hypervalent pentacoordinatedphosphorane. In
the case of phosphine oxides, addition of organometallic reagents
(e.g., alkoxides, Grignard reagents, and organolithium reagents) can
provide ready access to hypervalent pentacoordinatedphosphoranes.[96−98] For example, Newkome and Hager reported that heating phenyldi(2-pyridyl)phosphineoxides at 100 °C in the presence of EtONa afforded 2,2′-bipyridines
in 50–60% yield.[98] Phosphonium saltscontaining two 2-pyridyl groups display similar behavior. Oae and
Uchida demonstrated that treatment of benzyltri(2-pyridyl)phosphonium
bromide in water or alcohol in the presence or absence of acid could
result in ligand coupling.[102b−102e] They also found that when tri(2-pyridyl)phosphineoxide was treated with nucleophilic organometallic reagents, both
the coupling products of the two pyridine ligands and the coupling
products of the pyridine ligand and the incoming nucleophile were
formed.[102b] When using benzyl magnesium
chloride as the nucleophile, 2-benzylpyridine was formed as the major
product.[102b] Despite these advances, ligand
coupling of heterocycles at the phosphorus center has found limited
use because no general and mild conditions have been established to
access related phosphonium salts bearing heteroaryls.
In the
pioneering study reported by the Anders group, class="Chemical">pyridine
was transformed into a class="Chemical">pan class="Chemical">pyridylphosphonium salt in a site-selective
and efficient manner through sequential addition of Tf2O, PPh3, and Et3N.[104] In one of the limited examples of the application of this finding,
they showed that addition of NaN3 as the nucleophile to
the pyridylphosphonium salt resulted in coupling of the azido and
pyridine groups, followed by a Staudinger reaction to give the iminophosphorane.[104a] The potential of this methodology in ligand
coupling remained unexplored until recently, when McNally and co-workers
investigated the site-selective functionalization of pyridines (Scheme ).[99a] They found that, in most cases, the original
organic base (Et3N) gave good results when forming phosphonium
salts, but DBU was superior in some situations. The reaction temperature
was also a critical factor of this protocol, with some substrates
requiring low temperatures and some others needing higher temperatures.
The purification of the phosphonium salts was performed through simple
precipitation, without chromatography. The transformation displayed
good regioselectivity, with the 4-substituted product being the major
product. When the 4-position was blocked, 2-substituted products formed.
Various functionalities were tolerated; specifically, pyridines with
halogen substituents were introduced smoothly into the phosphonium
salt. Addition of sodium alkoxide to the phosphonium salt produced
the alkoxylated pyridine. Based on previous reports,[96−98] McNally and co-workers proposed the formation of the pentavalent
alkoxyphosphorane intermediate 6–5, followed by contractive ligand coupling to construct the C–O
bond. They could not, however, rule out the nucleophilic aromatic
substitution (SNAr) pathway. The reactions displayed good
functional group tolerance and site-selectivity. For example, both
nicotine and loratadine were transformed into phosphonium salts and
reacted with sodium alkoxide to afford the alkoxylated products in
good yields. Furthermore, sodium thiolate, sodium azide, and 2-pyridinelithium
were suitable nucleophiles for this phosphorus ligand coupling reaction.
Most recently, Vilotijevic and co-workers reported that benzothiazoles
can also be functionalized through a phosphorus(V) ligand coupling
process.[99h] Although Anders and co-workers
had reported the preparation of benzothiazol-2-yl-triphenylphosphonium
triflates previously, they did not fully study the scope and application
of these salts.[104b]
Scheme 42
Selective Functionalization
of Pyridines via Heterocyclic Phosphonium
Salts
In 2018, McNally and class="Chemical">co-workers
used class="Chemical">pan class="Chemical">phosphonium salts as a platform
for the synthesis of heterobiaryls through contractive C–C
coupling via P(V) intermediates (Scheme ).[99c] Facing
the challenge of preparing di(azaaryl)phosphonium salts under mild
conditions, they developed a unique route, without using heteroaryl
halides or a lithium reagent. The key to the introduction of two heterocycles
to the phosphorus center was the use of the fragmentable phosphine 6–8. Thus, the phosphonium triflate 6–9 underwent deprotonation by DBU, accompanied
by the loss of methyl acrylate, which then provided the heteroaryl
phosphine 6–10, positioning the first
coupling partner in place. Upon engagement with a second coupling
partner, the dihetarylphosphonium salt 6–11 formed with complete regiochemical control. Finally, an
acidic alcohol solution initiated phosphorus ligand coupling to provide
the heterobiaryl product 6–12. Numerous
examples of the reaction proceeded with synthetically useful overall
yields (17–34% over three steps). In addition, the synthesis
of complex heterobiaryls is possible, including the use of bioactive
molecules as coupling partners. The McNally group has also provided
further advances in phosphorus(V) ligand coupling.[99d−99g]
Scheme 43
Synthesis of Heterobiaryls through P(V) Contractive Coupling
Although the phenomenon of class="Chemical">phosphorus(V) ligand
class="Chemical">pan class="Chemical">coupling was observed
as early as 1948,[102a] its potential in
heterocycle functionalization was unveiled only recently. Contractive
phosphorus(V) coupling is a coupling method for heterocycle functionalization
that is complementary to transition-metal-mediated reactions in terms
of selectivity and functional group tolerance. In transition-metal-catalyzed
cross-coupling, redox cycling between the n and n + 2 oxidation states, through oxidative addition and reductive
elimination, is crucial for catalysis. The phosphoruscontractive
coupling resembles reductive elimination, albeit in a noniterative
manner. In the three-stage coupling sequence designed by McNally and
co-workers, they realized a combination of stoichiometric oxidative
addition and reductive elimination. In light of recent progress in
the field of main group element redox catalysis, one cannot help but
look forward to inventions that will enable P(III)/P(V) redox catalysis
for phosphorus ligand coupling.[101]
Conclusion and Outlook
class="Chemical">Phosphorus-based organocatalysis
is a vast field. This brief Overview
of various toclass="Chemical">pics in class="Chemical">pan class="Chemical">phosphorus-based catalysis highlights key recent
advances in the field, with emphasis on four topics: nucleophilic
phosphine catalysis, phosphine oxide catalysis, phosphorus-mediated
SET processes, and phosphorus(V) ligand coupling. Below, we provide
a summary and outlook for each topic.
In this well-established but still
active area, the present focus of nucleophilicclass="Chemical">phosphine catalysis
lies in the develoclass="Chemical">pment of new class="Chemical">patterns of annulations, along with
asymmetric variants of known reactions. Unforeseen reactions class="Chemical">pan class="Chemical">continue
to be discovered when employing highly engineered substrates. The
application of nucleophilic catalysis in the synthesis of functional
molecules remains promising. Moreover, nucleophilic addition of phosphines
to electron-deficient C–C multiple bonds promotes the in situ
formation of zwitterionic adducts that can act as efficient base catalysts
for various asymmetric transformations. With continued progress, we
expect to see the different reactivity patterns of nucleophilic phosphine
catalysis merging with transition metal catalysis and photoredox catalysis.
Reactions driven by class="Chemical">phosphineclass="Chemical">pan class="Chemical">oxide
formation remain highly popular among synthetic chemists. Catalytic,
economical, and environmentally benign versions of these reactions
will undoubtedly have a huge impact on synthesis in both academic
and industrial settings. Recent advances have unveiled insights into
how to improve the in situ activation of strong P(V)=O bonds.
Multiple studies have revealed that a strained phosphacycle scaffold
can facilitate the in situ reduction of phosphine oxides during redox-driven
catalysis. Denton and co-workers found that a carefully designed phosphineoxide catalyst could realize additive-free phosphine oxide catalysis
in a redox-neutral setting. Collectively, these findings signify that
the design of phosphine (oxide)s with appropriate skeletons will be
pivotal in moving the field forward. As demonstrated by Sevov and
co-workers, new technologies, including electrochemistry and photoredox
chemistry, can also add impetus to the in situ recycling of phosphineoxides. Alongside these advances, asymmetric phosphine oxide catalysis
is also appealing for preparing highly valuable chiral molecules.
Ultimately, we wish to see catalytic versions of traditional phosphineoxide-generating reactions applied in both academia and industry.
At present, organicclass="Chemical">phosphorusclass="Chemical">pan class="Chemical">compound-mediated
SET processes remain undeveloped. Although limited in scope, this
area is complementary to nucleophilic phosphine catalysis in terms
of its mechanisms and functional group tolerance. While nucleophilic
phosphine catalysis has been limited to electron-deficient multiple
bonds, phosphorus radical catalysis can activate inert alkenes and
C–H bonds. The design of phosphines or their complexes with
structural frameworks of suitable electronic potential may lead to
the open-shell species having longer half-lives, allowing rapid advances
in this area. Several recent accounts have revealed that the EDAcomplexes
of phosphines with RfI, NaI, or Cs2CO3 have a good SET ability, especially in the presence of visible light,
potentially inspiring more discoveries in this field. Phosphorus-based
reagents also facilitate the generation of active radical intermediates
through deoxygenation or desulfurization, especially under photoredox
catalysis. For the latter, challenges remain in the requirement for
stoichiometric phosphorus reagent and the generation of waste. These
issues may be addressed through recycling of the phosphine oxide,
as we have discussed in Section . With recent developments in photochemistry and electrochemistry,
we envision phosphoruscompound-mediated SET processes to be a promising
area.
class="Chemical">Phosphorus ligand
class="Chemical">pan class="Chemical">coupling displays
the versatility of phosphoruscompounds, and it can serve as a complementary
method for transition-metal-mediated coupling for heterocyclic functionalization.
Although such reactions remain stoichiometric at present, their reductive
elimination steps emulate those of transition-metal-promoted reactions.
Together with other main group elements (e.g., S, I, Bi), the realization
of catalytic P(V) ligand coupling may be on the horizon.
Finally, several adclass="Disease">ditional toclass="Chemical">pics—class="Chemical">pan class="Chemical">phosphonium
salt PTCs, iminophosphorane super base catalysis, asymmetric phosphineoxide Lewis base catalysis, FLP catalysis, and chiral phosphoric acid
catalysis—are also important aspects of phosphorus-based organocatalysis.
With the rich chemistry of organophosphoruscompounds, this area continues
to progress with the development of novel transformations. Phosphorus-based
catalysis is at the threshold of a bright future. With continued invention,
we anticipate phosphorus-based catalysis to have many more applications
in the chemical and pharmaceutical industries.
Authors: Emma E Coyle; Bryan J Doonan; Andrew J Holohan; Killian A Walsh; Florie Lavigne; Elizabeth H Krenske; Christopher J O'Brien Journal: Angew Chem Int Ed Engl Date: 2014-09-22 Impact factor: 15.336
Authors: Trevor V Nykaza; Gen Li; Junyu Yang; Michael R Luzung; Alexander T Radosevich Journal: Angew Chem Int Ed Engl Date: 2020-01-29 Impact factor: 15.336
Authors: Gen Li; Trevor V Nykaza; Julian C Cooper; Antonio Ramirez; Michael R Luzung; Alexander T Radosevich Journal: J Am Chem Soc Date: 2020-03-25 Impact factor: 15.419
Scheme 1
Scheme 2
Scheme 3
Scheme 4
Scheme 5
Scheme 6
Scheme 7
Scheme 8
Scheme 9
Scheme 10
Scheme 11
Scheme 12
Scheme 13
Scheme 14
Scheme 15
Scheme 16
Scheme 17
Scheme 18
Scheme 19
Scheme 20
Scheme 21
Scheme 22
Scheme 23
Scheme 24
Scheme 25
Scheme 26
Scheme 27
Scheme 28
Scheme 29
Scheme 30
Scheme 31
Scheme 32
Scheme 33
Scheme 34
Scheme 35
Scheme 36
Scheme 37
Scheme 38
Scheme 39
Scheme 40
Scheme 41
Scheme 42
Scheme 43