Download PDF
Synlett 2019; 30(17): 1954-1965
DOI: 10.1055/s-0039-1689924
cluster
© Georg Thieme Verlag Stuttgart · New York

Development of a Hydrazine-Catalyzed Carbonyl-Olefin Metathesis Reaction

Tristan H. Lambert  *
a   Department of Chemistry, Columbia University, New York, New York 10027, USA
b   Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, USA   Email: Tristan.lambert@cornell.edu
› Author Affiliations

The hydrazine catalysis research was funded by the National Science Foundation (CHE-0953259).
Further Information

Publication History

Received: 31 January 2019

Accepted after revision: 10 May 2019

Publication Date:
05 June 2019 (online)

 
 PDF Download Permissions and Reprints All articles of this category


This manuscript is dedicated to Professor Tom Katz, Columbia University, for his contributions to olefin metathesis.

Published as part of the Cluster Metathesis beyond Olefins

Abstract

Carbonyl-olefin metathesis is a potentially powerful yet underexplored reaction in organic synthesis. In recent years, however, this situation has begun to change, most notably with the introduction of several different catalytic technologies. The development of one of those new strategies, based on hydrazine catalysts and a novel [3+2] paradigm for double bond metathesis, is discussed herein. First, the stage is set with a description of some potential applications of carbonyl-olefin metathesis and a discussion of alternative strategies for this intriguing reaction.

1 Introduction

2 Potential Applications of Carbonyl-Olefin Metathesis

3 Carbonyl-Olefin Metathesis Strategies

4 Direct (Type I): Non-Catalytic

5 Direct (Type I): Acid-Catalyzed

6 Indirect (Type II): Metal Alkylidenes

7 Indirect (Type III): Hydrazine-Catalyzed

8 Conclusion


#

Key words

carbonyl-olefin metathesis - hydrazines - 1,3-dipolar cycloadditions - azomethine imines - cycloreversion

Biographical Sketch

Zoom Image

Tristan Lambert was born in Madison, WI, in 1976, and grew up in the small town of Black Earth. He graduated from the University of Wisconsin at Platteville in 1998 with a B.S. in chemistry. He then undertook his graduate work at UC-Berkeley (M.S. 2000) and at Caltech (Ph.D. 2004). He was subsequently an NIH postdoctoral fellow at the Memorial Sloan Kettering Cancer Center. In 2006, Tristan accepted a faculty position at the Department of Chemistry at Columbia University. In 2011 he was promoted to Associate Professor and in 2016 to Full Professor. In January 2018, he moved to the Department of Chemistry and Chemical Biology at Cornell University.

1

Introduction

Catalytic olefin metathesis has revolutionized how chemists approach the synthesis of complex molecules because it has enabled structural reorganizations that are difficult to replicate by other means.[1] Although far less developed, other double bond metathesis reactions should also offer great advantages for molecular synthesis. Arguably the most powerful and useful of these alternatives is carbonyl-olefin metathesis (COM) (Scheme [1]).[2] Indeed, the metathetical exchange of alkenes and carbonyls implies a number of exceptionally useful applications (vide infra). However, despite this promise, such transformations have seen remarkably little use, due to the limitations of existing approaches to carbonyl-olefin metathesis. Only recently have catalytic carbonyl-olefin metathesis methods been developed, and even these technologies remain in their infancy.

Fortunately, this situation has begun to change with the introduction of catalysts and strategies that effect carbonyl-olefin metathesis reactions efficiently and with an expanding range of substrates. In this account, we discuss our own contribution to this area,[3] which concerns the use of hydrazine catalysts along with a novel [3+2] cycloaddition paradigm for double-bond metathesis. To set the stage, we first discuss in general terms some of the most exciting potential uses of catalytic carbonyl-olefin metathesis, and then recap some of the alternative approaches to this intriguing reaction class.

Zoom Image
Scheme 1 Carbonyl-olefin metathesis

# 2

Potential Applications of Carbonyl-Olefin Metathesis

Analogous to olefin metathesis, the realization of a generally competent carbonyl-olefin metathesis process would enable a wide range of useful transformations (Scheme [2]). Indeed, direct parallels can be drawn between the two; however, the incorporation of a carbonyl as one of the double bond components suggests a number of important conceptual and practical differences. At the present time, many of the potential applications of carbonyl-olefin metathesis are at best highly substrate-limited, and in many cases wholly unrealized.

Zoom Image
Scheme 2 Potential applications of carbonyl-olefin metathesis

Perhaps the most obvious applications of carbonyl-olefin metathesis reside in the arena of ring-opening and ring-closing metathesis. For ring-opening carbonyl-olefin metathesis (ROCOM), a cyclic olefin 1 is converted into an acyclic molecule 3 with differentiated functional groups (Scheme [2a]). Functional group orthogonality is of course a very useful principle for synthesis, and it represents one of the most striking distinctions between COM and olefin metathesis. The reverse process, ring-closing carbonyl-olefin metathesis (RCCOM), represents another powerful application (Scheme [2b]). Because in this case the products are the same as those that would result from ring-closing olefin metathesis, the real advantage of RCCOM resides in the utilization of the carbonyl functionality leading up to the metathesis event. For example, one could readily imagine leveraging a variety of organocatalytic[4] technologies for the construction of a complex RCCOM substrate such as 4, which could then be converted into a cyclic olefin 5. Cross carbonyl-olefin metathesis (XCOM) for carbonyl olefination offers another intriguing potential application of this technology (Scheme [2c]). In this case, an aldehyde 6 or ketone would be subjected to an intermolecular COM reaction in the presence of an excess of olefin 7, thereby olefinating the carbonyl. In this way, XCOM would offer an attractive alternative to the Wittig[5] or Tebbe olefination[6] reactions, obviating the use of sensitive and non-atom-economical reagents. The converse process, in which an olefin 10 is subjected to XCOM in the presence of excess carbonyl 11 (Scheme [2d]), would result in the formal oxidative cleavage of the olefin and thereby provide an alternative to ozonolysis or strongly oxidizing metal reagents.[7] The combination of carbonyl olefination via XCOM and RCCOM offers yet another potential alternative to a classic reaction, the McMurry coupling (Scheme [2e]).[8] Here, a dicarbonyl 15 could conceivably be converted into a cyclic alkene 18 without resort to Ti(III) reagents or other such conditions. Finally, catalytic carbonyl-olefin metathesis could also be used to achieve olefin metathesis, thereby providing new possibilities for this venerable transformation (Scheme [2f]). To achieve such reactions, a catalytic amount of the carbonyl component 20 could be used, which after successive COM events, would achieve the same end (22) as direct olefin metathesis. This strategy raises the possibility, as yet unrealized, of developing an organocatalytic or Lewis acid catalyzed variant of the olefin metathesis reaction.

At the present time, these synthetic possibilities remain at best narrow in scope, and in many cases, undemonstrated. The full realization of these strategies will have to wait for the development of carbonyl-olefin metathesis technologies that are reactive and efficient enough to accommodate a much broader range of substrates. Encouragingly, work in this area has been increasing in recent years, and a number of important advances have been registered. Before discussing our group’s contribution to this area, we first discuss the state of the field. The discussion below is not intended to be a comprehensive treatment of carbonyl-olefin metathesis, but rather to provide an overview that contextualizes the thought process that went into our work.


# 3

Carbonyl-Olefin Metathesis Strategies

Approaches to carbonyl-olefin metathesis may be divided into three main categories, which we refer to here as types I–III (Scheme [3]). Type I represents the direct approach (Scheme [3a]), wherein the carbonyl and olefin components undergo cycloaddition to form an oxetane intermediate, followed by cycloreversion to furnish the metathesis products. The type I category includes both concerted, photochemical [2+2] cycloaddition reactions (the ‘Paterno–Büchi approach’), as well as acid catalysis as shown. Types II and III represent indirect approaches (Schemes 3b and 3c), in which one of the components, either the olefin (type II) or the carbonyl (type III), is converted into a new double-bonded species, which then engages the other component in a cycloaddition/cycloreversion process. The type II indirect approach encompasses the as yet unrealized idea of metal-alkylidene-promoted or catalyzed carbonyl-olefin metathesis. Meanwhile, the hydrazine-catalyzed approach we have described falls into the type III indirect category. Clearly, the opportunities and challenges for catalysis are significantly different with each of these approaches.

Zoom Image
Scheme 3 Catalytic strategies for carbonyl-olefin metathesis

# 4

Direct (Type I): Non-Catalytic

The oldest and arguably most conceptually straightforward approach to carbonyl-olefin metathesis is via the direct cycloaddition/cycloreversion of the carbonyl and olefin components. For the most part, this approach entails the use of photochemical [2+2] cycloaddition, a process known as the Paterno–Büchi reaction,[9] followed by heat- or acid-induced cycloreversion. For example, Scharf and Korte reported in 1963 that the bicyclic oxetane 50 underwent acid-induced cycloreversion to produce the alkenyl aldehyde 51 (Scheme [4a]).[10] Since the bicycle 50 is prepared by photocycloaddition of norbornene (48) and benzophenone (49), this process represents a straightforward, albeit stepwise, approach to carbonyl-olefin metathesis (although the transformation was not described as such).

Grigg reported in 1972 that [Rh(CO)2Cl]2 catalyzed this same cycloreversion under rather mild conditions (Scheme [4b]).[11] It is most probable that the rhodium catalyst was acting simply as a Lewis acid, inducing stepwise cycloreversion via initial rupture of the bisbenzylic C–O bond. It can be imagined that this catalytic cycloreversion could be conducted in tandem with the photocycloaddition to produce bicycle 50, thereby realizing a one-step process, but this combination has not yet been reported.

Zoom Image
Scheme 4 ROCOM of norbornene via a Paterno–Büchi reaction followed by acid-mediated cycloreversion

Although these previous examples clearly accomplished the same goal, it was not until 1973 that Jones first labeled this two-step process as a carbonyl-olefin metathesis.[12] In this work, a range of examples was demonstrated in which the cycloreversion step was induced by pyrolysis (Scheme [5]).

Zoom Image
Scheme 5 Thermal oxetane cycloreversion to furnish the carbonyl-olefin metathesis product

Jones followed up this study in 1975 with the ring-opening carbonyl-olefin metathesis of cyclohexene (55) with benzaldehyde (56) and the demonstration that, in addition to pyrolysis, the cycloreversion of intermediate 57 could be promoted by either a Brønsted acid or a rhodium complex with much better olefin geometry selectivity for the formation of 58 (Scheme [6]).[13]

Zoom Image
Scheme 6 ROCOM via a Paterno–Büchi reaction followed by acid-mediated cycloreversion

Meanwhile, Carless showed that cycloreversion of oxetanes 61, accessed from benzaldehyde derivatives 59 and olefin 60, could be promoted by more conventional Lewis acids including BF3·OEt2 and AlCl3, although only with selected substrates (Scheme [7]).[14] Specifically, it was shown that substrates bearing electron-neutral or electron-deficient aryl rings led to exclusive ring expansion/rearrangement products 62 rather than the desired metathesis products 63. Only with electron-rich aryl rings was metathesis the major pathway.

Zoom Image
Scheme 7 XCOM via a Paterno–Büchi reaction followed by Lewis acid mediated fragmentation

Despite its potential complications, this two-step approach can be useful for rapid developments in molecular complexity. As a prime example, Kutateladze has shown the transformation of 64 into the carbonyl-olefin metathesis product 66 by way of oxetane 65 (Scheme [8]).[15] This example demonstrates that when properly constrained, oxetane ruptures can be relatively well-behaved.

Zoom Image
Scheme 8 Complex rearrangements via Paterno–Büchi reaction/fragmentation sequences

Not surprisingly, certain oxetane intermediates are prone to spontaneous cycloreversion, typically when a substituent is present that offers significant stabilization of a latent carbocation. For instance, D’Auria demonstrated that the photocycloaddition of furan carboxaldehyde 67 with furan (68) leads to the formation of bicyclic oxetane 69, which spontaneously converts into dienal 70 (Scheme [9a]).[16] Interestingly, the cycloadduct 72 derived from dihydrofuran 71 and benzaldehyde (56) has been shown to convert into 73, the ring-opened carbonyl-olefin metathesis product, under photoirradiation in the presence of a photosensitizer (Scheme [9b]).[17]

Zoom Image
Scheme 9 ROCOM of furan or dihydrofuran via Paterno–Büchi reactions

The virtues of the photochemical approach to carbonyl-olefin metathesis include the relative simplicity of the reaction conditions, at least in theory, and the direct engagement of substrates. Unfortunately, the need for substrates bearing chromophores, the general need for high energy radiation, and the practical drawbacks of running photochemical reactions stand as major impediments to the generalization of this approach. Furthermore, the difficulties of achieving catalyst or reagent control in the photochemical approach complicate the issues of realizing high regio- and stereoselectivity. Alternative approaches are thus of great interest.

A related approach to the Paterno–Büchi strategy for carbonyl-olefin metathesis is represented by those transformations that involve the direct formation and destruction of an oxetane intermediate, but without recourse to photochemical promotion. Such transformations are possible with specialized substrates that are capable of undergoing stepwise [2+2] cycloadditions, that is to say, substrates capable of facile carbocation formation, typically under the action of a Brønsted or Lewis acid.

An early example of this type of carbonyl-olefin metathesis was reported by Snider, which involved the treatment of dienyl ketone 74 with a mixture of methylaluminum dichloride and dimethylaluminum chloride to produce the tricyclic oxetane 75 (Scheme [10]).[18] Lewis acid promoted fragmentation of 75 then led to the bicyclic diene 76 in low yield.

Zoom Image
Scheme 10 Lewis acid mediated RCCOM

Bickelhaupt demonstrated the olefination of benzaldehyde (56) with 1,1-disubstituted or trisubstituted olefins (e.g. 77) to produce β-substituted styrenes 79 in low yield (Scheme [11]).[19] Under the acidic clay promoted conditions, acetone (78) liberated from the metathesis consumed an equivalent of benzaldehyde in an aldol condensation reaction, which underlines a potential problem with approaches of this type that are too forcing.

Zoom Image
Scheme 11 Acidic clay promoted XCOM

A ring-closing carbonyl-olefin metathesis promoted by BF3·OEt2 was developed by Schmalz for the synthesis of indene 81 in remarkably high yield (Scheme [12]).[20] Undoubtedly, the success of this reaction is due in large part to the involvement of tertiary and bisbenzylic carbocations as the key intermediates.

Zoom Image
Scheme 12 Lewis acid mediated RCCOM of a diaryl ketone

One of the most exceptional examples of this strategy, notable as much for the complexity of the substrate as for the completely aliphatic nature of the reactive functionalities, is the transformation of steroidal ketone 82 into 83 reported by Khripach (Scheme [13]).[21] Although not explicitly described as a carbonyl-olefin metathesis reaction, this example nicely demonstrates some of the potential utility that this reaction offers. As a personal aside, this reaction was the original inspiration for our own efforts in this area.

Zoom Image
Scheme 13 Lewis acid promoted rearrangement via carbonyl-olefin metathesis

The acid-promoted, stepwise version of the type I carbonyl-olefin metathesis has the potential benefit of offering external control over the course of the reaction. For obvious reasons, it would be much more attractive if the acid reagent could be used in catalytic quantities. Indeed, this approach has been shown to be viable in just the past several years.


# 5

Direct (Type I): Acid-C atalyzed

Given the above examples of Lewis or Brønsted acid mediated COM reactions, it seemed plausible that catalytic variants should work with appropriate substrates. Yet it was only recently that catalytic acid-catalyzed COM reactions were demonstrated. Franzén was able to demonstrate such a process between benzaldehyde (56) and trisubstituted alkenes 77 using trityl ion 84 as the catalyst (Scheme [14]).[22] Although the substrates are heavily biased to support the carbocation intermediates required by the reaction pathway and the catalyst loading was quite high (20 mol%), this work demonstrated that the acid-catalyzed approach could be efficient. More recently, Franzen has extended trityl catalysis to a RCCOM reaction to form indenes.[23]

Zoom Image
Scheme 14 Trityl cation catalyzed XCOM

Some of the most prominent work in this area has been reported by Schindler,[24] who showed in 2016 that the RCCOM of alkenyl ketones (e.g. 85 to 86) could be accomplished with high efficiency using iron(III) chloride as the catalyst (Scheme [15a]). This work was impressive for the mild conditions (room temperature) and reasonable catalyst loadings (5–10 mol%) needed for high yields, as well as the substrate scope which went significantly beyond previous work. A screen of different catalysts revealed that, although other Lewis acids were viable, superior efficiency was achieved with iron(III) chloride. Schindler has subsequently extended this strategy to the synthesis of polycyclic aromatics such as 88 (Scheme [15b]).[25]

Zoom Image
Scheme 15 Iron(III) chloride catalyzed RCCOM reactions

Shortly after the initial disclosure by Schindler, Li reported a similar method using iron(III) chloride catalysis (Scheme [16]).[26] Although similar in many respects, this work demonstrated the formation of six-membered rings as well as nitrogen heterocycles (e.g. 91), which represents a significant expansion of functionality for the area of COM. Schindler has also reported a dihydropyrrole synthesis method using RCCOM.[27]

Zoom Image
Scheme 16 Iron(III) chloride catalyzed RCCOM to form a dihydropyrrole

Although iron(III) chloride seems to be the best catalyst for these reactions, a variety of acids have been shown to catalyze similar processes (Scheme [17]). In addition to the trityl catalysis shown by Franzen and discussed above, Tiefenbacher has found that supramolecularly bound hydrochloric acid can serve as a viable COM catalyst,[28] while Nguyen has shown that tropylium ion (93)[29] and even molecular iodine[30] can catalyze COM reactions on suitably disposed substrates.

Zoom Image
Scheme 17 Alternative Lewis or Brønsted acid catalysts for carbonyl-olefin metathesis

Meanwhile, Schindler has recently reported that ring-opening carbonyl-metathesis is possible using gallium(III) chloride as the Lewis acid (Scheme [18]),[31] which was found to be superior to other catalysts. Although the yields were generally modest, this work represents an impressive extension of the Lewis acid catalysis strategy to the ROCOM arena.

Zoom Image
Scheme 18 Lewis acid catalyzed ROCOM

An interesting alternative carbonyl-olefin metathesis strategy involving photoredox catalysis was recently described by Glorius (Scheme [19]).[32] In this case, a styrene 97 and an aldehyde 96 underwent conversion into the intermediate diol 98 in the presence of an acridinium photocatalyst, water, formic acid, and irradiation with visible light. This intermediate was then subjected to an acid-catalyzed fragmentation reaction at elevated temperature to reveal the metathesized olefin 99 and acetone (78). Although this two-step process cannot be properly considered a catalytic carbonyl-olefin metathesis reaction, it is a useful example of how clever reaction design can offer new inroads in the field of carbonyl-olefin metathesis.

Zoom Image
Scheme 19 Two-step carbonyl-olefin metathesis involving photocatalysis and acid-mediated fragmentation

Given the versatility and tunability of Lewis and Brønsted acids, it seems likely that this approach to COM will continue to bear fruit. On the other hand, it also seems likely that this approach will be constrained to substrates that can accommodate carbocationic intermediates as required by the stepwise [2+2] pathway. Hence the search for a COM catalyst to rival olefin metathesis catalysts in terms of generality is probably going to require an alternative strategy altogether. Such was our thinking when we began to approach this problem in 2007. Before discussing those efforts, however, a discussion of another potential approach to carbonyl-olefin metathesis using metal alkylidenes is warranted.


# 6

Indirect (Type II): Metal Alkylidenes

Given the profound successes of olefin metathesis methods, it is extremely tempting to consider employing metal alkylidenes for the metathesis of olefins and carbonyls. In fact, carbonyl olefinations with metal alkylidenes are well known and have been put to impressive use for the total synthesis of highly complex natural products.[33] Unfortunately, achieving catalysis with this strategy has not yet proved possible, and the foreseeable challenges to doing so are significant.[34] Although they have been referred to as such,[2] these transformations are not carbonyl-olefin metathesis reactions, since a new carbonyl is not generated (Scheme [20]). (They are a class of double bond metathesis reaction, however.) Nevertheless, stoichiometric metal alkylidene-mediated carbonyl olefination methods offer a unique and powerful tool for complex molecule construction, and a review of these methods is both instructive and enlightening.

Zoom Image
Scheme 20 Stylized carbonyl olefination with metal alkylidenes

One of the earliest examples of metal-mediated carbonyl-olefin metathesis was reported by Jossifov, who found that tungsten(VI) chloride polymerized α,β-unsaturated ketones such as 104 or 106 to polyenes 105 or 107 with the excision of benzaldehyde or acetone respectively (Scheme [21]).[35] Although described as a ‘catalyst’, the WCl6 is added in superstoichiometric amounts, and there is little reason to expect that turnover of the tungsten oxo product was achieved under these conditions.

Zoom Image
Scheme 21 Polyene synthesis via carbonyl olefination

Grubbs demonstrated the utility of metal alkylidene-mediated olefin metathesis/carbonyl-olefination for complex molecule synthesis with the preparation of capnellene (Scheme [22]).[36] Treatment of tricycle 108 with the Tebbe reagent produced titanacyclobutane 109, which when heated to 90 °C underwent conversion into cyclobutene 111, presumably via the [2+2] cycloreversion of intermediate 110. Tricycle 111, shown in an alternative view 112, was subsequently carried forward to capnellene 113. Although not technically a carbonyl-olefin metathesis since no organic carbonyl was generated, this transformation clearly has close analogy to the type of process that might be envisioned with a true carbonyl-olefin metathesis reaction.

Zoom Image
Scheme 22 Total synthesis of capnellene

Grubbs and Fu subsequently demonstrated the synthesis of cycloalkenes from alkenones by tandem olefin metathesis/carbonyl olefination (Scheme [23]).[37] For example, ketone 114, when treated with 1 equivalent of the molybdenum alkylidene 116, produced cycloheptene 115 in only 30 minutes. Key to the success of this transformation was the fact that 114 engages in olefin metathesis much more readily than it does carbonyl-olefination, as the reverse sequence would have led to an acyclic diene. As noted above, for this process to be rendered catalytic it would be necessary to identify conditions under which a viable molybdenum alkylidene species could be regenerated from the Mo=O byproduct that results from olefination.

Zoom Image
Scheme 23 Intramolecular carbonyl olefination

Despite the fact that the metal alkylidene strategy has not yet been rendered catalytic, it has been put to further notable use for complex molecule synthesis. For example, Nicolaou utilized this strategy to high advantage as part of a fragment coupling strategy in the synthesis of complex polyethers (Scheme [24]).[38] Treatment of the ester 117 with 4 equivalents of the Tebbe reagent effected ring closure to produce the pentacyclic compound 118 in 71% yield. One of the clearest advantages to this approach is the fact that it allowed two bicyclic components to be merged by way of a simple esterification reaction.

Zoom Image
Scheme 24 Complex polyether synthesis via intramolecular carbonyl-olefin ring closure
Zoom Image
Scheme 25 Cyclic vinyl ether synthesis via intramolecular carbonyl-olefin ring closure

Ranier has also utilized titanium alkylidenes for the ring-closing olefination of alkenyl esters such as substrate 119 (Scheme [25]).[39] In this case, the precise nature of the alkylidene used was crucial to the selective formation of the cyclic enol ether 120 versus the acyclic enol ether (the classic ‘Tebbe’ product). In the absence of a strategy to recycle the titanium oxide byproduct, this synthetically useful process is necessarily a stoichiometric one.


# 7

Indirect (Type III): Hydrazine-Catalyzed

The preceding discussion serves to highlight some of the challenges associated with the development of a general catalytic platform for carbonyl-olefin metathesis. Specifically, the direct photochemical approach is not mediated by an external reagent, and catalysis via the metal alkylidene strategy requires the invention of a means to overcome a high thermodynamic barrier for catalyst turnover. At the time we were first contemplating a solution to this problem (2007), the direct acid-mediated approach was severely limited in scope. Recent successes with this approach notwithstanding, it seemed unlikely to us that a process involving stepwise [2+2] cycloadditions would offer a general enough approach to broadly realize the applications described above. We thus felt that a fundamentally different approach would provide the best opportunity to circumvent these formidable challenges. However, by rejecting the above strategies, we were left with the daunting question of what such an approach might be.

Zoom Image
Scheme 26 Traditional and alternative double bond metathesis paradigms

To answer this question, we took a step back and asked a different question: What is fundamentally occurring in a double bond metathesis reaction? As shown schematically in Scheme [26a], existing double bond metathesis reactions were understood to operate by the merger of two double-bond components via a [2+2] cycloaddition (concerted or stepwise) to produce a locally symmetric, four-membered intermediate. This intermediate would then undergo a retro [2+2] cycloaddition in an orthogonal sense to produce two new double bonded species, i.e., the metathesis products. The key insight for us was recognizing that what was essential about such a process was not that it occurred via a [2+2] pathway, but rather that the intermediate possessed the proper local symmetry to allow cycloreversion to occur in a manner orthogonal to the pathway that formed the cycloadduct. Crucially, we realized the requirement of local symmetry could be met by alternative reaction manifolds, perhaps most obviously those arising from [3+2] cycloadditions (Scheme [26b]).[40] Such a process would benefit from relying on well-established, thermally allowed pericyclic processes and would offer a unique means to engage carbonyl substrates, namely via their conversion into 1,3-dipolar reactants.

Among the numerous possibilities suggested by Scheme [26b], we decided that the most attractive opportunity lay in the application of azomethine imine 1,3-dipolar cycloadditions.[41] Our catalytic design for this strategy is shown in Scheme [27]. Thus a 1,2-disubstituted hydrazine 121 can condense with an aldehyde 122 to generate an azomethine imine 123, which can then participate in 1,3-dipolar cycloadditions with olefins 124 to produce pyrazolidines 125. Importantly, these pyrazolidines satisfy the local symmetry design described above. Cycloreversion of the pyrazolidine 125 would then reveal the new olefin 126 and the azomethine imine 127. Finally, hydrolysis of 127 would furnish the new aldehyde 128 and regenerate the hydrazine catalyst 121. Although this process is inherently thermoneutral, the equilibrium could be biased by release of ring strain, conjugation, or the use of mass action, just as with traditional olefin metathesis.

Zoom Image
Scheme 27 Generic catalytic cycle for hydrazine-catalyzed carbonyl-olefin metathesis

Before describing the implementation of this strategy, it is important to point out that there existed strong precedent for each of the steps in this catalytic cycle. The most well-supported aspect of this design was of course the azomethine imine cycloaddition reaction, which was pioneered by Huisgen in the 1960s (Scheme [28a]).[42] A wealth of subsequent work by Huisgen, Oppolzer,[43]and numerous other researchers has established this 1,3-dipolar cycloaddition as a reliable means to access substituted pyrazolidines. Although the majority of these reactions employ azomethine imines with an electron-withdrawing substituent on one of the nitrogen atoms, examples exist in which unactivated dipoles are utilized. A particularly cogent example for our study was the reaction reported by Shimizu shown in Scheme [28b], in which a 1,2-disubstituted hydrazine 132 participates in an azomethine imine 1,3-dipolar cycloaddition to produce the tricycle 133 in quantitative yield.[44] Interestingly, Shimizu also reported that 1,1-dimethylhydrazine (135) also participates in 1,3-dipolar cycloaddition with benzaldehyde (56) and styrene (134) to produce the pyrazolidinium product 136 (Scheme [28c]), demonstrating that the dipole need not be charge neutral.

Zoom Image
Scheme 28 Selected examples of 1,3-dipolar cycloadditions of azomethine imines/hydrazonium ions

Crucially, though few in number, examples also exist of the retrocycloaddition of pyrazolidines.[45] [46] [47] For example, it was shown that cycloreversion of the ketone 137 could be accomplished by heating in xylenes over extended periods of time (Scheme [29a]).[46b] The resulting azomethine imine 138 was trapped by cycloaddition with norbornene (48) to furnish adduct 140 in 52% yield. This work demonstrated that conjugation of the newly formed olefin helped to accelerate the cycloreversion, as the corresponding ketal of substrate 137 did not undergo reaction under these conditions.

Zoom Image
Scheme 29 Examples of pyrazolidine cycloreversions

The effect of conjugation could also be enhanced, as in the case of 141 reported by Hamelin (Scheme [29b]), which underwent cycloreversion in refluxing toluene to generate benzylidene malonate 143 along with azomethine imine 142.[46c] The latter was trapped with dimethyl fumarate (144) to furnish the new cycloadduct 145. Unfortunately, no time or yields were given for this reaction.

Another intriguing example of pyrazolidine cycloreversion, and the one most relevant to our work described below, is that in a report by Zettl in 1977 of a ‘dipole metathesis’ of azomethine imine 146 (Scheme [29c]).[46a] Here the 1,3-dipole 146 engaged in cycloaddition/cycloreversion with a variety of electron-deficient alkenes, such as tetracyanoethylene (147), which led to the formation of the new azomethine imine 148 and diene 149. Although in these reactions the original dipole was not derived from a carbonyl compound, nor was one formed afterwards, this work demonstrated the essence of the carbonyl-olefin metathesis strategy that we had envisioned.

With this background in mind, we decided to investigate the viability of the carbonyl-olefin metathesis design described in Scheme [27], by targeting the ring-opening of cyclopropenes (Scheme [30]). Due to their high ring strain,[48] we anticipated that these substrates would facilitate both cycloaddition and cycloreversion and thus give us the best chance for success. After some experimentation, we found that the bicyclic hydrazine[49] catalyst 47·2HCl was able to effect ROCOM of cyclopropene 150 with benzaldehyde (56) to produce the alkenal 151 in 84% yield over 24 hours at 75 °C in DCE. Importantly, benzaldehyde and cyclopropene 150 were not observed to undergo any reaction in the absence of catalyst, nor in the presence of only HCl. In addition to the reaction shown, we found that hydrazine 47·2HCl efficiently catalyzed the metathesis of a number of other cyclopropenes and aldehydes, a selection of which are shown in Scheme [30].

Zoom Image
Scheme 30 Hydrazine-catalyzed ROCOM of cyclopropenes

The mechanistic rationale for this metathesis process closely follows the one described in Scheme [27], except with the intermediacy of hydrazoniums 159/160 instead of azomethine imines (Scheme [31]), due to the fact that the catalyst 47 is used as its dihydrochloride salt.[50] Cycloaddition of 159 with cyclopropene 161 produces pyrazolidinium salt 162,[51] which after proton transfer (162 to 163), undergoes cycloreversion to produce hydrazonium ion 164. Hydrolysis of 164 then furnishes the metathesis aldehyde 165 with concomitant regeneration of hydrazine catalyst 47.

Zoom Image
Scheme 31 Proposed catalytic cycle for hydrazine-catalyzed ROCOM

Several lines of evidence supported this rationale. First, we were able to prepare the hydrazonium perchlorate 166,[50] a stable and crystalline solid that corresponded to our proposed reactive intermediate (Scheme [32]). This species was found to undergo stoichiometric metathesis with cyclopropene 150 to produce the aldehyde 151. In addition, hydrazonium 166 and the ring-opened hydrazonium 167 were identified by 1H NMR in the catalytic reaction. Finally, theoretical calculations by Houk fully supported this mechanistic hypothesis and helped to quantify the activation energies and free energy changes for each step in this process.[3b]

Zoom Image
Scheme 32 Evidence for the reaction pathway via stoichiometric reaction

While illuminating, these calculations also underscored the challenges faced in extending the substrate scope of this catalysis. By design, the cyclopropane ring-strain enabled the pericyclic events to occur at relatively modest temperatures. Unfortunately, less-strained olefins were calculated to have activation energies for cycloreversion that were significantly higher. Indeed, although we avoided some of the difficulties of the [2+2] pathway, we created a new problem in the form of the highly stable pyrazolidine ring.

Fortunately, the design of this catalytic approach is such that the hydrazine structure provides massive opportunity for structural variation. Based on fundamental principles, it stands to reason that lower energy barriers and thus more broadly applicable catalysis should be possible. In addition, the participation of only first-row elements makes computational screening of catalyst structure a reasonably straightforward effort. Indeed, such efforts are currently underway in our lab, and we have identified structures that expand hydrazine-catalyzed carbonyl-olefin metathesis in a significant way. The first fruits of these efforts will be reported shortly.


# 8

Conclusion

As described earlier, catalytic carbonyl-olefin metathesis holds significant promise for enabling a wide variety of useful new strategies for organic synthesis. Although the barriers to realizing processes that are as efficient and general as olefin metathesis are substantial, the increasing dedication of effort in this area bodes well that such solutions are on the horizon. It is encouraging that such a diversity of approaches has proven viable, if not yet general. It seems likely that the maturation of carbonyl-olefin metathesis into a robust technology for synthesis will require a combination of these different strategies.


#
#

   Permissions and Reprints All articles of this category
Zoom Image
Zoom Image
Scheme 1 Carbonyl-olefin metathesis
Zoom Image
Scheme 2 Potential applications of carbonyl-olefin metathesis
Zoom Image
Scheme 3 Catalytic strategies for carbonyl-olefin metathesis
Zoom Image
Scheme 4 ROCOM of norbornene via a Paterno–Büchi reaction followed by acid-mediated cycloreversion
Zoom Image
Scheme 5 Thermal oxetane cycloreversion to furnish the carbonyl-olefin metathesis product
Zoom Image
Scheme 6 ROCOM via a Paterno–Büchi reaction followed by acid-mediated cycloreversion
Zoom Image
Scheme 7 XCOM via a Paterno–Büchi reaction followed by Lewis acid mediated fragmentation
Zoom Image
Scheme 8 Complex rearrangements via Paterno–Büchi reaction/fragmentation sequences
Zoom Image
Scheme 9 ROCOM of furan or dihydrofuran via Paterno–Büchi reactions
Zoom Image
Scheme 10 Lewis acid mediated RCCOM
Zoom Image
Scheme 11 Acidic clay promoted XCOM
Zoom Image
Scheme 12 Lewis acid mediated RCCOM of a diaryl ketone
Zoom Image
Scheme 13 Lewis acid promoted rearrangement via carbonyl-olefin metathesis
Zoom Image
Scheme 14 Trityl cation catalyzed XCOM
Zoom Image
Scheme 15 Iron(III) chloride catalyzed RCCOM reactions
Zoom Image
Scheme 16 Iron(III) chloride catalyzed RCCOM to form a dihydropyrrole
Zoom Image
Scheme 17 Alternative Lewis or Brønsted acid catalysts for carbonyl-olefin metathesis
Zoom Image
Scheme 18 Lewis acid catalyzed ROCOM
Zoom Image
Scheme 19 Two-step carbonyl-olefin metathesis involving photocatalysis and acid-mediated fragmentation
Zoom Image
Scheme 20 Stylized carbonyl olefination with metal alkylidenes
Zoom Image
Scheme 21 Polyene synthesis via carbonyl olefination
Zoom Image
Scheme 22 Total synthesis of capnellene
Zoom Image
Scheme 23 Intramolecular carbonyl olefination
Zoom Image
Scheme 24 Complex polyether synthesis via intramolecular carbonyl-olefin ring closure
Zoom Image
Scheme 25 Cyclic vinyl ether synthesis via intramolecular carbonyl-olefin ring closure
Zoom Image
Scheme 26 Traditional and alternative double bond metathesis paradigms
Zoom Image
Scheme 27 Generic catalytic cycle for hydrazine-catalyzed carbonyl-olefin metathesis
Zoom Image
Scheme 28 Selected examples of 1,3-dipolar cycloadditions of azomethine imines/hydrazonium ions
Zoom Image
Scheme 29 Examples of pyrazolidine cycloreversions
Zoom Image
Scheme 30 Hydrazine-catalyzed ROCOM of cyclopropenes
Zoom Image
Scheme 31 Proposed catalytic cycle for hydrazine-catalyzed ROCOM
Zoom Image
Scheme 32 Evidence for the reaction pathway via stoichiometric reaction