DOI: 10.1002/asia.200900712

Transition-Metal-Catalyzed Cycloadditions for the Synthesis of Eight-Membered Carbocycles

Zhi-Xiang Yu,* Yi Wang, and Yuanyuan Wang[a]

Dedicated to the 100th anniversary of the College of Chemistry, Peking University

1072 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWS

Abstract: Eight-membered carbocycles are found in awide variety of natural products that exhibit a broadrange of biological and medicinal activities (cf. the mostpotent anticancer drug, taxol). However, the synthesis ofeight-membered carbocycles is quite challenging as tradi-tional approaches are met with entropic and enthalpicpenalties in the ring-forming transition states. These nega-tive effects can be totally or partially avoided with the im-plementation of transition-metal-catalyzed/mediated cy-cloadditions. In this Focus Review, examples of elegant

and efficient metal-catalyzed and some metal-mediatedcycloadditions (including Ni- catalyzed [4+4] and Rh-cat-alyzed [4+2+2] and [5+2+1] reactions) are presented toillustrate this. Application of these cycloaddition reactionsin total synthesis is also presented to show the significanceof these reactions in addressing challenges in naturalproduct synthesis.

Keywords: cycloaddition · eight-membered carbocycles ·total synthesis · transition metals

Introduction

Carbocycles of various ring sizes are widely found in naturalproducts and often incorporated into drugs. While five- andsix-membered carbocycles are most common, seven- andeight-membered or larger carbocycles occupy a significantfraction of these compounds.[1] Figure 1 shows some naturalproducts containing eight-membered carbocycles. Althoughthe number of natural products containing eight-memberedcarbocycles is limited, these natural products have garneredthe attention of synthetic organic and medicinal chemists.[1,2]

This interest is a result of the medicinal and biological prop-erties these compounds possess, making them attractive asdrugs, drug leads, or biological probes.[3] Taxol, the mostwell-known natural product in this family, is now among themost potent anticancer drugs in clinical use.[4] In 2008, an-other member of this family—pleuromutilin—was approvedfor use as an anitbiotic by the FDA.[5]

Another draw to study eight-membered carbocyclescomes from the general challenge of synthesizing medium-sized ring systems (usually 8- to 12-membered rings). Tradi-tional ring-closure strategies are effective for small ring sys-tems, but problematic for medium-sized ring systems owingto entropic and transannular penalties incurred when bring-ing the two ends of a reactant together.[6]

To date, several methods to form eight-membered carbo-cycles have been developed including ring-closing metathe-sis and Cope rearrangement.[7] Despite these important ad-vances, there is a lack of generality to accommodate manytarget structures.[2c] Considering the complexity of naturalproducts containing eight-membered carbocycles, new reac-tions that are both general and robust would be readily wel-comed. In recent years, significant progress in this direction

[a] Prof. Dr. Z.-X. Yu, Dr. Y. Wang, Dr. Y. WangBeijing National Laboratory for Molecular Sciences (BNLMS)Key Laboratory of Bioorganic Chemistry and Molecular Engineeringof the Ministry of EducationCollege of ChemistryPeking UniversityBeijing 100871 (China)Fax: (+86) 10-6275-1708E-mail : yuzx@pku.edu.cn

Figure 1. Representative natural products with eight-membered carbocy-clic skeletons.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1073

has been made through the de-velopment of transition-metal-catalyzed cycloadditions—thefocus of this review.

The mechanism of metal-catalyzed cycloadditions allowsfor efficient ring formation ofmedium-sized ring compounds.These reactions usually pro-ceed by coordination, oxida-tive coupling, insertion, and re-ductive elimination processes,as schematically presented bythe three-component transition-metal-catalyzed [m+n+o]cycloaddition shown in Figure 2. Therefore, entropic and en-

thalpic penalties associated with the traditional ring-closuremethods could be circumvented or partially avoided in themetal-catalyzed cycloadditions.

On the other hand, some two-component cycloadditionssuch as [4+4] and [6+2] reactions to synthesize eight-mem-bered carbocycles are forbidden according to the Wood-ward–Hoffmann rules. However, transition-metal-catalyzedcycloadditions can achieve these reactions via the mecha-nisms shown in Figure 2, providing another impetus to dis-cover metal-catalyzed cycloadditions for the synthesis ofeight-membered carbocycles.

In 1996, Lautens and co-workers summarized manymetal-catalyzed cycloadditions including metal-catalyzed cy-cloadditions for eight-membered carbocycle synthesis.[2b]

Now 14 years have passed since Lautens� review, and manynew results, which will be covered in this Focus Review,have appeared in the literature during this period. To givean overview of all these achievements, several importantand representative examples given in Lautens� and others�reviews[2] have also been included here. In the presentreview, we pay specific attention to metal catalysis; however,several stoichiometric metal-mediated cycloadditions toeight-membered carbocycles have also been described. Inaddition, we have included the proposed mechanisms of sev-eral recently discovered cycloadditions.

In what follows, we can see that cycloadditions take placein either an intramolecular or an intermolecular fashion.They could be two-component, three-component, or multi-component. All these cycloadditions are traditionally namedas [m+n], [m+n+o], [m+n+o+x] cycloadditions. Here m, n,o, and x are the numbers of carbon units which are incorpo-rated in the eight-membered carbocycles. Here we wish tointroduce a new nomenclature for these cycloaddition reac-tions. If a reaction is a two-, three-, or four-component reac-tion, this reaction is called an [m+n], [m+n+o], or[m+n+o+p] reaction. In this case, m, n, o, and p compo-nents come from different molecules. If a reaction is writtenas a [(m+n+o)] reaction, this implies that it is an intramo-lecular reaction, where m, n, and o units come from onemolecule. In a similar way, an [(m+n)+o] reaction meansthat it is a two-component intermolecular reaction, where mand n units come from one molecule, but the o unit comesfrom another molecule. Compared to the traditional nomen-clature, the present nomenclature is not easy, but can pro-

Zhi-Xiang Yu was born in Ezhou, Hubeiprovince, China in 1969. He obtained hisBS (Wuhan University, 1991), MS (syn-thetic organic chemistry, Peking Universi-ty, 1997, with Prof. Qingzhong Zhou),PhD (computational and theoretic organ-ic chemistry, the Hong Kong University ofScience & Technology, 2001, with Prof.Yun-Dong Wu). After a three-year post-doc study (UCLA, with Profs. K. N.Houk (2001–2004) and M. Masca (2001–2002)), he joined Peking University as anassociate professor in 2004 and was pro-moted to a full professor in 2008. His

main research interests are to apply computational organic chemistry andsynthetic organic chemistry to study reaction mechanisms, develop new re-actions and catalysts, and synthesize natural and non-natural products. Re-cently he won the National Science Fund for Distinguished Young Schol-ars of China in 2008.

Yi Wang was born in Chongqing, China.She obtained her BS degree in ChongqingThree Gorges College in 2004, and herPhD from Sichuan University in 2009under the supervision of Prof. Xiang-GeZhou. Now she is a postdoctorate re-searcher with Prof. Zhi-Xiang Yu atPeking University. Her research interestsinclude asymmetric catalysis, transition-metal catalytic reactions, computationalchemistry, and total synthesis of naturalproducts.

Yuanyuan Wang was born in Tianjin,China. She obtained her BS degree fromWuhan University in 2004, and her PhDfrom Peking University in 2009 under thesupervision of Prof. Zhi-Xiang Yu. Nowshe is an industrial research scientist.

Figure 2. Depiction of the process of transition-metal-catalyzed [m+n+o] cycloaddition.

1074 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

vide more information to the referred reactions. For exam-ple, this nomenclature can tell us how a reaction really hap-pens and where these carbon units come from. We prefer touse this new nomenclature for every discussed cycloadditionreaction below. When we collect similar reactions togetherfor a subtitle in this review, we will still use the traditionalnomenclature of metal-catalyzed cycloadditions.

1. ACHTUNGTRENNUNG[2+2+2+2]Cycloadditions

The first [2+2+2+2] cycload-dition of acetylene to form cy-clooctatetraene (COT) wasdiscovered by Reppe in thelate 1940s. The originally re-ported tetramerization of acet-ylene was catalyzed by NiBr2/CaC2 (Scheme 1). Later on itwas found that [Ni ACHTUNGTRENNUNG(acac)2] or [Ni ACHTUNGTRENNUNG(cot)2] can also catalyzethis reaction. This [2+2+2+2] reaction has now been ap-plied in the synthesis of COT in industry.[8]

The above [2+2+2+2] reaction led to a mixture of vari-ous substituted COT compounds when substituted alkyneswere used. To overcome this shortcoming, tom Dieck[9] de-veloped a regioselective [2+2+2+2] cycloaddition of mono-substituted alkynes to generate 1,3,5,7- or 1,2,4,6-tetrasubsti-tuted cyclooctatetraenes (Scheme 2). The catalyst used byhis group was a 1,4-diazadiene nickel complex, a Ni catalystwith two bulkyl ligands.

A solution to control the regioselectivity of alkyne tetra-merization was to use diyne substrates, which can undergotwo-component [(2+2)+ ACHTUNGTRENNUNG(2+2)] cycloaddition to give COTin the presence of Ni catalysts. However, this tetrameriza-tion reaction often suffers from a competitive (and domi-

nant in most cases) [(2+2)+2] cyclotrimerization reaction,leading to substituted benzenes. In 2007, Wender[10,11a] andco-workers found that the two-component [(2+2)+ACHTUNGTRENNUNG(2+2)]cycloaddition reaction of diynes with various tethers becamefavored dramatically over the cyclotrimerization by using[NiBr2ACHTUNGTRENNUNG(dme)] and zinc dust in THF at 60 8C (Scheme 3). Thecatalytic species was believed to be Ni0, in situ generated by

the reduction of [NiBr2 ACHTUNGTRENNUNG(dme)] with zinc. It was found thatthe efficiency and selectivity of the [(2+2)+ACHTUNGTRENNUNG(2+2)] reactionwas affected by the catalyst loading and catalyst concentra-tion. Increasing the catalyst loading from 20 to 40 mol % im-proved the selectivity of COT over arene from 6.9:1 to 52:1.The Ni-catalyzed [(2+2)+ ACHTUNGTRENNUNG(2+2)] reaction was scalable toobtain the COT compounds on a gram scale and can be ex-tended to synthesize unsymmetrical COTs if two differentdiynes were used. In 2009, Wender and co-workers furtherdemonstrated that the [(2+2)+ACHTUNGTRENNUNG(2+2)] reaction can be car-ried out by using mixed diynes, with both internal and termi-nal yne units, providing hexa- and octasubstituted COTs ingood yields (Scheme 3).[11b] When aromatic or heteroaromat-ic diynes were used as substrates, the ratio of COT/benzenederivatives was as high as 20:1. Recently, Okamoto and co-workers extended Wender�s [(2+2)+ ACHTUNGTRENNUNG(2+2)] reaction proto-col by using ionic liquid supported nickel complexes in anionic liquid/toluene biphasic system.[11c]

One drawback of the above mentioned Ni-catalyzed[(2+2)+ ACHTUNGTRENNUNG(2+2)] cycloaddition with diynes was that a largeamount of the Ni catalyst (20–40 mol %) had to be used. In2009, Zhao and co-workers found that 1,6-enynes can alsoundergo the [(2+2)+ACHTUNGTRENNUNG(2+2)] cycloaddition to give 1,5-cyclo-octadienes together with some [(2+2)+2] products whenlower loading of Ni/R2Zn (4 mol %) was used as the catalyst(Scheme 4).[12] It is interesting to note that the choice of a

Scheme 1.

Scheme 2.

Scheme 3.

Scheme 4.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1075

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

combination of different Ni complexes and R2Zn had a dra-matic influence on the reaction of 1,6-enynes. When a [Ni-ACHTUNGTRENNUNG(acac)2]/Me2Zn combination was employed, high total yieldswith poor selectivities ([(2+2)+ACHTUNGTRENNUNG(2+2)]/ ACHTUNGTRENNUNG[(2+2)+2]=1:1)were observed for carbon-tethered aryl enynes (here R= Arin Scheme 4). Unidentifiable polymeric products or a com-plex reaction mixture were obtained using N-tosyl and N-benzyl enynes as substrates under these catalytic conditions.However, for the [NiCl2 ACHTUNGTRENNUNG(PCy3)2]/Me2Zn combination, thesesubstrates invariably favored the formation of the [(2+2)+2]products in moderate yields with ratios of [(2+2)+2] cyclo-adducts over [(2+2)+ACHTUNGTRENNUNG(2+2)] cycloadducts ranging from 4:1to 7:1. In contrast, when Et2Zn instead of Me2Zn, or only0.6 equivalents of Et2Zn, was used as the reducing reagentunder the same reaction conditions, reductive eneyne cycli-zation products were exclusively formed.

2. ACHTUNGTRENNUNG[4+4] Cycloadditions

In the early 1950s, Ziegler and Reed discovered that buta-diene can dimerize and oligomerize with a metal catalyst(Scheme 5).[13] The dimerization reactions led to two prod-

ucts, cyclooctadiene (via a [4+4] reaction), and 4-vinyl cy-clohexene (via a [4+2] reaction). After optimization of thereaction conditions, Wilke and Heimbach discovered thatthe reaction yield of the [4+4] cycloaddition to give cyclooc-tadiene (COD) can reach 95 % when a phosphine ligandwas introduced to the Ni reaction system.[14] The yield wasfound to be dependent on both the ligands and the ratio ofcatalyst to ligand. When a bis ACHTUNGTRENNUNG(p-allyl) nickel complex wasused, cyclooctadiene 9 was formed in preference to otherproducts such as vinylcyclohexene (VCH) 10, divinylcylobu-tane, and cyclic trimers of butadiene.

Stimulated by the seminal studies shown in Scheme 5,Waegell and co-workers developed a regioselective [4+4]cycloaddition of monosubstituted butadienes in 1985(Scheme 6).[15] The final cycloadducts were generated by ahead-to-head cycloaddition mode, and the substituents in

the final product have a trans configuration. tom Dieck[16]

found that an iron complex can also catalyze this reactionand they demonstrated that the asymmetric version of thisreaction can be achieved if a chiral catalyst is used(Scheme 7). In 1995, Itoh and co-workers found that

[(C5R5)Ru ACHTUNGTRENNUNG(diene)]+ can catalyze/mediate dimerization ofconjugated dienes via a [4+4] cycloaddition.[17] However,preliminary results showed that the regioselectivity of Itoh�s[4+4] cycloaddition was not satisfactory.

Considering the common existence of eight-memberedcarbocycles fused with other rings in natural products,Wender[18] and co-workers sought to explore an intramolecu-lar [(4+4)] reaction that could have application in naturalproduct synthesis. They succeeded in developing [(4+4)] cy-cloadditions to synthesize 5/8 and 6/8 bicyclic COD(Scheme 8). In the 5/8 system, when 13 was treated in tolu-ene at 60 8C with 11 mol % [NiACHTUNGTRENNUNG(cod)2] and 33 mol % PPh3,the major product is cis-fused, whereas for the 6/8 system,trans products dominate (15 to 16). A type-II Ni-catalyzed[(4+4)] reaction was also developed by Wender�s group tosynthesize bridge-fused bicyclic cyclooctadienes (17 to18).[19]

Scheme 5.

Scheme 6.

Scheme 7.

Scheme 8.

1076 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

A total synthesis of the naturally occurring sesquiterpeneasteriscanolide using Ni-catalyzed intramolecular [(4+4)] cy-cloaddition was achieved by Wender.[20] The eight-mem-bered ring-forming transformation of 19 to 20, the key stepof this total synthesis, was accomplished in 67 % yield andgave a single diastereomer (Scheme 9).[20a] Wender also uti-

lized this [(4+4)] reaction to construct the 5/8/5 skeleton ofophiobolin F (Scheme 10).[20b] The key cycloaddition was ac-complished by treatment of bis-diene 21 with 10 mol % [Ni-ACHTUNGTRENNUNG(cod)2] and 20 mol% triphenylphosphine in toluene at 60 8C,followed by desilylation, affording bicyclic cyclooctadiene 22in 60 % yield. After further elaborations from 22, the corestructure of ophiobolin F was constructed.

In 1999, Murakami and Ito reported a palladium-cata-lyzed [4+4] cycloaddition reaction of vinylallenes to con-struct eight-membered carbocyclic rings (Scheme 11).[21]

They found that the R1 group in 23 must be either a phenylor vinyl group for this [4+4] reaction. Otherwise, only [4+2]cycloadditions took place. In some cases, the [4+4] reactionsare competitive with the intermolecular [4+2] cycloadditionsof vinylallenes.

The above Pd-catalyzed [4+4] cycloaddition was furtherdeveloped by Lee and co-workers.[22] In 2006, they reported

a Pd-catalyzed intermolecular [4+4] cycloaddition reactionof vinylallenes, generated in situ from a-bromovinyl arenesand propargyl bromides (Scheme 12). The formed [4+4] cy-cloadduct 27 can act as a diene to further react with various

dienophiles to give a 6/8 bicyclic skeleton (for example,compound 28 in Scheme 12), allowing rapid synthesis ofeight-membered carbocycles with excellent yields. More im-portantly, they found that this reaction could be accom-plished in one pot to achieve high synthetic efficiency.

A possible reaction pathwayfor the reaction in Scheme 12is described as follows(Scheme 13). Oxidative addi-tion of a Pd0 catalyst to thevinyl bromide gives 29. Thentransmetalation between 29and propagyl bromide/indiumcomplex, followed by reductiveelimination, leads to vinylal-lene 26. Oxidative cyclometal-

lation of 26 to Pd affords intermediate 30, which subse-quently reacts with another molecule of vinylallene at theinternal double bond of the allene to give a bis ACHTUNGTRENNUNG(s-allyl)palla-dium intermediate 31. Subsequent reductive elimination

Scheme 9.

Scheme 10.

Scheme 11.

Scheme 12.

Scheme 13.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1077

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

from 31 gives rise to eight-membered carbocycle 27. ThisPd-catalyzed [4+4] catalytic cycle of vinylallenes was origi-nally proposed by Murakami and Ito for the reaction shownin Scheme 11.

3. ACHTUNGTRENNUNG[4+2+2] Cycloadditions

The first [4+ACHTUNGTRENNUNG(2+2)] reaction was reported in 1970 by Carbo-naro,[23a] who discovered that [FeACHTUNGTRENNUNG(acac)3] can catalyze the re-action between 1,3-butadiene and norbornadiene (NBD) togive [4+ACHTUNGTRENNUNG(2+2)] cycloaddition product 32 in 25 % yield and[2+ ACHTUNGTRENNUNG(2+2)] adduct 33 in 12 % yield, together with someNBD dimers (Scheme 14). Later on Carbonaro, Inukai, and

Lyons found that cobalt complexes can catalyze the [4+

(2+2)] cycloadditions of NBD with dienes also, usually withbetter yields compared to those obtained by iron cata-lysts.[23b–d]

Lautens[24] and co-workers reported the asymmetric induc-tion of this [4+ACHTUNGTRENNUNG(2+2)] cycloaddition employing cobalt cata-lysts in the presence of a chiral phosphine ligand andEt2AlCl (as a reducing agent). [CoACHTUNGTRENNUNG(acac)2] and [CoACHTUNGTRENNUNG(acac)3]were both effective as precatalysts and gave similar levels ofasymmetric induction. Of the various chiral phosphinestested, (R)-Prophos gave the highest enantiomeric excess of70 % (Scheme 15).

Even though the above [4+2+2] reactions above sufferedfrom low reaction yields and produced excess NBD dimers,they provided inspirations for further development of[4+2+2] reactions to synthesize eight-membered carbocy-cles.

In 2002, Evans[25] and co-workers discovered a Rh-cata-lyzed two-component [4+ ACHTUNGTRENNUNG(2+2)] reaction of butadiene and

heteroatom-tethered enyne derivatives to synthesize eight-membered carbocycles (Scheme 16). The tethers can be ni-trogen, sulfur, or oxygen atoms. The silver salt was criticalto the success of Evans� [4+ACHTUNGTRENNUNG(2+2)] reaction. If AgOTf was

used to generate a cationic Rh catalyst from Wilkinson cata-lyst, high yields of the [4+ ACHTUNGTRENNUNG(2+2)] reaction can be achieved.However, if AgSbF6 was used, the dominant products werethe homodimers generated via a Rh-catalyzed [(2+2)+ACHTUNGTRENNUNG(2+2)] reaction of the enynes. Evans found that in the lattercase, butadiene underwent oligomerization and could nottake part in the [4+ACHTUNGTRENNUNG(2+2)] reaction. Consequently, the re-maining enyne in the reaction system would undergo homo-dimerization. This [(2+2)+ ACHTUNGTRENNUNG(2+2)] reaction of enynes cata-lyzed by Rh complements the same reaction catalyzed by Niin the Zhao group (see Scheme 4).[10]

Evans and co-workers found that when utilizing a rhodi-um N-heterocyclic carbene (NHC) complex, [RhACHTUNGTRENNUNG(COD)-ACHTUNGTRENNUNG(IMes)Cl], as the catalyst, it was very efficient to achievehigh diastereoselectivities in the [4+ ACHTUNGTRENNUNG(2+2)] reaction of sub-stituted enynes 38 and butadiene (Scheme 17).[26] This wasthe first example of a rhodium-NHC complex catalyzed[m+n+o]-type cyclization to give eight-membered carbocy-cles with high diastereoselectivity.

A sequential three-component allylic amination/[4+

(2+2)] reaction can be realized to synthesize bicyclic cyclo-octadienes (Scheme 18).[26] Treatment of allylcarbonate withthe lithium salt of N-tosylpropargylamine in the presence ofWilkinson�s catalyst, modified with silver triflate, gaveenyne 40, which was not isolated but directly subjected toheat at reflux for 12 h under an atmosphere of 1,3-buta-

Scheme 14.

Scheme 15.

Scheme 16.

Scheme 17.

1078 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

diene. In this case, the tandem reaction afforded the cyclo-addition adduct in 87 % yield.

In 2002, Gilbertson[27] explored a new Rh-catalyzed[(4+2)+2] cycloadditon of dienynes and terminal alkynes(Scheme 19). They found that the cycloaddition reaction cat-alyzed by [Rh ACHTUNGTRENNUNG(nbd)Cl]2 and AgSbF6 in the presence of Me-DuPhos at 60 8C gave the final cycloadducts in good to ex-

cellent yields. The alkyne portion of the dienynes was a ter-minal alkyne. If the R1 group in dienynes 42 was hydrogen,the [(4+2)+2] cycloadditions with terminal alkynes gave amixture of two regioisomers 44 and 45, owing to the two dif-ferent insertion modes of the terminal alkyne to the Rh�Cbond. However, if the R1 group was methyl, only one regio-isomer 44 was generated. Gilbertson found that in 44, theR1 group and the bridgehead hydrogen had a trans configu-ration (see Scheme 19). Simple acetylene can also take partin the [(4+2)+2] reaction with dienynes. It is interesting tonote that, although a chiral ligand was used in Gilberson�sreaction, no enantioselectivity was observed. Only one[(4+2)+2] reaction examined had an ee value of 41 %. Thiswas different from the Rh-catalyzed intramolecular [(4+2)]reaction of dienynes, wherein a high-chirality transfer fromthe ligand to the products was observed by Gilbertson.[28] Ifno external alkyne was added, dienynes underwent[(4+2)+2] reactions between two dienynes, where one dien-yne provided the diene (four-carbon component) and yne(two-carbon component) parts, and the other dienyne usedits yne part (two-carbon component).

Even though Evans and Gilbertson�s reactions can gener-ate eight-membered carbocycles, their reactions cannot tol-erate substrates with carbon tethers (with the exception ofcarbon tethers in the form of an aromatic ring). In 2006,Wender and co-workers reported a new [(4+2)+2] reactionbetween ene-dienes and terminal alkynes catalyzed by

[Rh(CO)2Cl]2 and AgSbF6 indichloroethane solvent(Scheme 20).[29] One of the sig-nificant features of Wender�s[(4+2)+2] reaction was thatcarbon, nitrogen, and oxygentethers can all be incorporated.Usually the Wender [(4+2)+2]cycloaddition gave a mixtureof two regioisomers using theterminal alkynes as two-carbonsynthons. It was found that theregioselectivity of the alkyneinsertion was influenced by

both steric and electronic features of the alkyne. Wenderand his co-workers further extended this two-component re-action into a three-component [4+2+2] reaction by using2,3-dimethyl-1,3-butadiene, norbornene, and methyl prop-argyl ether as substrates.

In 2006, Murakami[30] reported a nickel-catalyzed intermo-lecular [4+ACHTUNGTRENNUNG(2+2)] annulation reaction of cyclobutanoneswith diynes that furnished bicyclic eight-membered cyclicketones (Scheme 21). They found that the N-heterocycliccarbene ligand IPr improved the activity of the nickel cata-lyst compared to the phosphine ligand, making the reactionoccur at room temperature and have excellent reactionyields. When unsymmetrical diynes were used, a 4:1 mixtureof regiosomers 54 and 55 was obtained with PACHTUNGTRENNUNG(nBu)3 as theligand, whereas only 54 was obtained when the stericallybulkier IPr ligand was used. The use of unsymmetrical 2-substituted cyclobutanones was also examined, showing ahigh regioselectivity (>20:1) under Ni-IPr catalysis.

The authors postulated the following mechanism(Scheme 22). The diyne and cyclobutanone initially bind onnickel(0) to give complex 56. Then intermediate 57 a or 57 b,or both, is obtained through oxidative cyclometalation ofthe two alkyne moieties to nickel(0), or through hetero-typeoxidative cyclization of the carbonyl group. Subsequent in-corporation of the third unsaturated functionality into theNi�C bond of the five-membered nickelacycles leads to theformation of oxanickelacycloheptadiene 58. Then the four-

Scheme 18.

Scheme 19.

Scheme 20.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1079

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

membered ring of this spiro nickelacycle is opened by b-carbon elimination to expand the seven-membered nickel-acycle. Finally, reductive elimination gives the eight-mem-bered product and regenerates the catalyst of nickel(0).

In contrast to the nickel-catalyzed cyclotetramerization ofalkynes (see Scheme 1), cobalt-catalyzed reactions of al-kynes mainly lead to [2+2+2] cyclotrimerizations to formbenzene derivatives. In 2008,Hilt[31] et.al successfully cou-pled two alkynes with 1,3-dienes to generate eight-mem-bered ring systems in an unpre-cedented three-component[4+2+2] fashion (Scheme 23).In this reaction, they utilized acobalt catalyst in the presenceof Zn, Fe, and ZnI2. The

[4+2+2] products were formedregioselectively with the twosubstituents from the alkynesin a 1,2-relation. The additionof iron powder reduced theamount of side products andincreased the yield of theeight-membered ring products.When acceptor-substituted ter-minal alkynes reacted withelectron-rich 1,3-dienes, thecorresponding cyclooctadie-nones could be obtained in ex-cellent yields (up to 90 %).

Saa[32] and co-workers re-ported a Ru-catalyzed [4+

(2+2)] cycloaddition betweendiynes and 1,3-butadienes in2003 (Scheme 24). They found

that under the catalysis of 10 % Ru complex and 10 %Et4NCl in DMF, diynes and 1,4-substituted dienes gave [4+ACHTUNGTRENNUNG(2+2)] cycloadducts of 1,3,5-cyclooctatrienes and 1-vinyl-1,3-cyclohexadienes. This formal [4+ACHTUNGTRENNUNG(2+2)] cycloaddition

was a tandem process and required addition of Et4NCl tothe reaction system. The purpose of the addition of Et4NClwas to neutralize the positively charged moiety in [Cp*Ru-ACHTUNGTRENNUNG(CH3CN)3]PF6, generating a neutral complex of [Cp*RuII-ACHTUNGTRENNUNG(CH3CN)2Cl] in situ as the catalyst.

Saa proposed the following reaction mechanism to explaintheir experimental results (Scheme 25). Firstly, oxidative ad-dition of 1,6-diyne to the Ru catalyst gives a rutheniumcyclic species 64. Then diene insertion into the Ru�C bondin 64 gives rise to two seven-membered intermediates 65 aand 65 b. The ratio of 65 a/65 b was dependent on the substi-tution pattern in the butadiene component, and was deter-

Scheme 21.

Scheme 22.

Scheme 23.

Scheme 24.

1080 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

mined by the steric and electronic nature of the diene inser-tion transition states. 65 a and 65 b then undergo b-hydrideelimination and reductive elimination to give 66 and 67, re-spectively, together with regeneration of the catalyst. Finally,intermediate 66 undergoes 8p electrocyclization to furnishthe [4+ACHTUNGTRENNUNG(2+2)] product of 62, whereas intermediate 67 un-dergoes 6p electrocyclization to give vinylcyclohexadiene63.

4. ACHTUNGTRENNUNG[6+2] Cycloadditions

The first metal-catalyzed [6+2] cycloaddition was reportedby Pettit,[33] who showed that cycloheptatriene-iron tricar-bonyl complex 68 and dimethyl acetylenedicarbornate underphotolysis gave an iron-diene complex (Scheme 26). The

photolysis here aimed to release a CO molecule from cyclo-heptatriene-iron tricarbonyl complex to generate an activecycloheptatriene-iron dicarbonyl complex, which then un-derwent [6+2] reaction with electron deficient acetylene. Fi-nally CO complexation gave the tricarbonyl complex 69.Even though this [6+2] reaction had low reaction yield, thisoriginal discovery inspired further research in metal-mediat-ed/-catalyzed [6+2] cycloadditions.

In 1983, Mach[34] and co-workers reported a metal-cata-lyzed [6+2] cycloaddition for accessing eight-memberedrings. The cycloaddition reaction between cycloheptatriene(CHT) and butadiene in the presence of 1 mol% TiCl4 and

20 mol % Et2AlCl led to thecycloadduct 70 bearing aneight-membered ring with abicycloACHTUNGTRENNUNG[4.2.1] nonane frame-work (Scheme 27). Under theoptimized conditions, adduct70 a was obtained in 78 %yield. When butadiene was re-placed by isoprene, cycloaddi-tion occurred exclusively onthe less substituted olefin, pro-viding 70 b in 61 % yield. CHTcould also react with alkynesto give [6+2] cycloadducts 71.

Recently, Schmidt[35] report-ed a molybdenum complex cat-alyzed [6+2] addition between

pinocarvone and cycloheptatriene to give an eight-mem-bered cycloadduct in excellent yield (Scheme 28). Using mo-lybdenum h4-oxadiene complexes as the catalysts, the[6+2]cycloaddition reactions proceeded smoothly at 50 8C orbelow. But analogous tungsten complexes exhibited a slight-ly lower catalytic activity toward the [6+2] reaction.Green[36] also reported [6+2] additions of CHT and COTwith disubstituted acetylenes catalyzed by molybdenum h4-oxadiene complexes to give the corresponding cycloadducts74 and 75, respectively.

Kreiter[37] and co-workers reported their first study of the[6+2] reaction of [(cht)Cr(CO)3] complexes with variousdienes under photochemical conditions. They observed that[6+2] and [6+4] cycloadditions could occur, depending onthe structure of the dienes. But further study of the originsof the different regioselectivities and of how to control theregioselectivity has not been reported.

Later on, Rigby[38] and co-workers examined in detail thecycloaddition of Cr complexes with various types of alkeneswith regard to selectivity and synthetic potential. High dia-stereoselectivity was observed in the reaction of 76 a with

Scheme 25.

Scheme 26.

Scheme 27.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1081

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

ethyl acrylate to give 77 a in 92 % yield as a single diastereo-mer (Scheme 29). Reaction of a stereochemically homoge-neous chromium complex 76 b revealed that complexationof the alkene to the metal occurred prior to carbon–carbonbond formation, in accord with Pettit�s results obtainedfrom the iron complexes. The position of the substituent onthe cht complexes was critical to the regioselectivity of thiscycloaddition. For example, single regiosomers were ob-served with 1-substituted cht ligands. However, when 2-sub-stituted cht was used, no selectivity was observed, irrespec-tive of the electronic nature of the substituents on the cht li-gands. Surprisingly, the level of selectivity with 3-substitutedcht was excellent with respect to an electron-poor cht, and

modest with respect to an electron-rich cht. A catalyticamount of a naphthalenechromiun tricarbonyl complex caninduce cycloaddition when azepines participate as the trienepartners in this reaction. This catalytic [6+2] reaction gave82 (Scheme 29) with a reaction yield of 77 % and a catalystturnover number of larger than 8. The [6+2] reactions ofcht-Cr complex with allenes were developed by Rigby in2008.[38h] Interestingly, Rigby and co-workers developed aresin-based chromium catalyst that can perform [6+2] cyclo-additions of a variety of cht derivatives and a,b-unsaturatedesters at 150 8C.[38g]

Rigby designed a concise approach to synthesize cedreneby using the chromium-mediated [6+2] reaction(Scheme 30).[39] The [6+2] cycloaddition of 83 was conduct-

ed at 150 8C, furnishing tricyclic ester 84 in 80 % yield. Afterfurther elaboration, cedrene was synthesized successfullyfrom 84.ACHTUNGTRENNUNG[(cht)Cr(CO)3] complexes can react with alkenes to give[6+2] products; however, they fail to react with alkynes.Buono[40a] found that the [CoI2ACHTUNGTRENNUNG(dppe)]/Zn/ZnI2 system effec-tively catalyzed the [6+2] cycloaddition of CHT and termi-nal alkynes without irradiation (Scheme 31). When a chiralligand binol phosphoramidite was used, the enantioselective

cycloaddition of CHT and phe-nylacetylene afforded the cy-cloadduct in up to 91 % yieldand 74 % ee. Importantly, thecatalyst system could tolerate awide range of functionalgroups such as ketone, sulfone,ester, alcohol, imide, and ni-trile. In 2008, Buono devel-oped this Co-catalyzed [6+2]reaction to its asymmetric ver-sion by introducing chiralphosphoramidites as ligands(up to 92 % ee was achie-ved).[40b] In 2005, Buono andco-workers found that both cy-clooctatetraene and 1,3,5-cy-clooctatriene react with termi-nal alkynes to give [6+2] cyclo-adducts under the colbalt cata-lysis.[40c] In 2009, Hilt and co-workers made further advancein this field, showing that CHTreacted with both internal al-

Scheme 28.

Scheme 29.

Scheme 30.

1082 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

kynes and terminal alkenes to give [6+2] cycloadducts inthe presence of catalytic amounts of cobalt dibromide [bis-(triisopropylphosphite) comlexes.[40d]

Buono and co-workers proposed the following mechanismfor the Cr-catalyzed [6+2] reaction (Scheme 32). Firstly

[CoL2I2] is reduced by zinc metal and ZnI2 to a cationic[CoL2]

+ complex. Then coordination of the alkyne and cy-cloheptatriene to the metal center, followed by oxidativecyclization, lead to the generation of cobaltacyclopentene86. A series of 1,5-migrations of the C ACHTUNGTRENNUNG(sp3)�Co bondhappen through consecutive s,p-allyl complexes 87, 88, and89. Finally reductive elimination from 89 generates the[6+2] cycloadduct 85 and the catalyst.

An intramolecular [(6+2)] cycloaddition between CHTand an alkyne to afford cycloheptatrienes was developed re-cently by Tenaglia[41] and co-workers using PtCl2 as the cata-lyst (Scheme 33). Complex heterocyclic compounds wereobtained when this reaction was applied to substrates con-taining a heteroatom (O or N) in the tether. A proposedmechanism accounting for the [(6+2)] cycloaddition is givenin Scheme 33.

Wender[42] reported a Rh-catalyzed [(6+2)] cycloadditionbetween vinylcyclobutanones and alkenes in 2000(Scheme 34). This reaction proceeded in good to excellentyields for substrates with N, C, and O tethers. High stereose-lectivity was found in this reaction in most cases, where cis-fused products were formed exclusively or preferentially.Quaternary carbon centers can be installed at the bridge-head carbon atoms of the final [(6+2)] cycloadducts. In ad-dition, this [(6+2)] reaction can be extended to an intramo-

lecular [(6+2)] reaction between vinylcyclobutanones andallenes.

Rodr�guez-Otero[43] studied the mechanism of the Wender[(6+2)] cycloaddition using density functional theory calcu-lations in 2008 (Scheme 35). It was proposed that this reac-

tion starts with oxidative addition of 94 to the catalytic spe-cies of [Rh(CO)Cl],[44] giving metallacyclopentane inter-mediate 96. Then b-carbon elimination leads to a nine-mem-bered-ring rhodacycle 97. Finally reductive elimination fur-nishes the [(6+2)] cycloadduct 95 and regenerates thecatalyst.

In 2006, Kuninobu and Takai[45a] reported a rhenium-cata-lyzed two-component [m+2]-like reaction between 1,3-ketoesters and alkynes, where m can be 5, 6, 7, or 8 (Scheme 36).

Scheme 31.

Scheme 32.

Scheme 33.

Scheme 34.

Scheme 35.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1083

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

This is an especially efficient[6+2] reaction for synthesizingeight-membered rings. The for-mation of 99 proceeded effi-ciently in toluene at 50 8C for24 h by using 5 mol %[ReBr(CO)5] as the catalystand 5 mol% benzyl isocyanideas the addiditive. Treatment of98 with arylacetylenes bearing substituents such as electron-donating methoxy and methyl groups and electron-with-drawing trifluoromethyl and bromo groups afforded the cor-responding eight-membered cyclic compounds in 86–97 %yields. In 2009, they found that the reaction system can bepromoted by the addition of 4 � molecular sieves instead ofa catalytic amount of phenylacetonitrile.[45b]

This Re-catalyzed addition was proposed to start with oxi-dative addition of the enol ether tautomer and the terminalacetylene (Scheme 37). Then a sequence of reactions involv-ing ring opening, retro-aldol reaction, isomerization and re-ductive elimination give the final [m+2] products.

5. ACHTUNGTRENNUNG[5+2+1] Cycloadditions

In 2001, Wender[46] and co-workers reported the first transi-tion-metal-catalyzed intermolecular [5+2+1] cycloadditionreactions of activated alkynes, CO, and vinylcyclopropanes(VCPs), leading to eight-membered rings (Scheme 38). The

final products obtained from Wender�s [5+2+1] reactionwere bicycloACHTUNGTRENNUNG[3.3.0]octenones 105, which were generated,during the hydrolytic workup, through the intramolecularaldol reaction of the in situ generated [5+2+1] cycloadducts104. In some cases, a small amount of [5+2+1] reactionproducts 104 was obtained. The alkynes used in Wender�s[5+2+1] reaction were carbonyl activated. Very high regio-

selectivities could be realized when unsymmetric activatedalkynes were used, showing that the R2 group, not the car-bonyl groups (R1 groups), preferred to be adjacent to the in-serted CO group in the final products.

Wender[47] and co-workers found that allene can also actas the two-carbon unit to take part in the three-component[5+2+1] reaction (Scheme 39). Under standard reactionconditions (1 mol % [Rh(CO)2Cl]2, dioxane as solvent, oneatmosphere of CO gas), cyclooctanedione 107 and its trans-annular aldol product hydroxybicycloACHTUNGTRENNUNG[3.3.0] octanone 108were obtained in 88 % yield in only 1 h. Increasing the COpressure from 1 atm to 2 atm improved the total yield to94 % and modestly changed the ratio of 107/108 from 1:2.2to 1:1.3, but required a longer reaction time of 15 h. Bothproducts were obtained as single E isomers. Here the usedallene was not carbonyl substituted, suggesting that allene ismore reactive than alkyne in this [5+2+1] reaction.

In 2007, Yu[48] and co-workers reported a two-component[(5+2)+1] reaction of ene-vinylcyclopropanes (ene-VCPs)and CO to give bicyclic cyclooctenones 110 (Scheme 40).This reaction was efficient with low CO pressure and thesubstrates can have carbon, nitrogen, and oxygen tethers.The [(5+2)+1] reaction gave only cis-5/8 bicyclic cycloocte-nones when R3 was hydrogen. However, when R3 was amethyl group, the formed 5/8 bicyclic cyclooctenones can bein a cis (when cis ene-VCPs were used) or trans (when transene-VCPs were used) configuration. The [(5+2)+1] reactioncan also be used to synthesize 6/8 bicyclic compounds, inwhich trans-fused 6/8 compounds were dominant.

This [(5+2)+1] reaction was also effective to reach 5/8/5systems if the cyclopropane ring of the ene-VCP was fusedby a five-membered carbocycle (Scheme 41).[49] It was foundthat, in the final products, the bridgehead C2 and C3 hydro-gen atoms always had a cis configuration. The configurationof the bridgehead C1 was also cis with respect to C2 and C3when R was hydrogen. However, when R was a methylgroup, the final 5/8/5 compounds were a mixture of both cisand trans configurations at the C1 atom (with respect to theC2 atom, see numbering of the carbon atoms in Scheme 41).

Scheme 36.

Scheme 37.

Scheme 38.

1084 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

The proposed mechanism for the [(5+2)+1] reaction isshown in Scheme 42. The catalytic cycle starts with theligand exchange of the product-catalyst complex, 5/8-ring-fused cyclooctenone-Rh(CO)Cl complex 119, with the ene-VCP substrate, followed by VCP coordination and openingof the cyclopropyl group, giving 115. Then coordination andinsertion of the alkene leads to rhodabicyclooctene 116. COcoordination and insertion affords rhodabicyclononenone118, which subsequently undergoes a migratory reductiveelimination to give the [(5+2)+1] cycloadduct and com-pletes the [(5+2)+1] catalytic cycle.

Application of this [(5+2)+1] cycloaddition to the synthe-ses of natural products containing eight-membered carbocy-clic skeletons and linear and branched triquinane skeletonswas demonstrated by Yu and co-workers.[50a] A very efficienttandem RhI-catalyzed [(5+2)+1]/aldol cycloaddition processwas developed to construct the triquinane skeleton and thisstrategy was successfully ap-plied to the total syntheses of(� )-hirsutene and (� )-desoxy-hypnophilin with high stepeconomy (Scheme 43).

Later on, a stepwise strategyto synthesize hirsutene was re-alized by the Yu group. In thisstrategy, the [(5+2)+1] cyclo-adduct 123 was transformed tosymmetric diketone 124. Thena sequence of aldol reaction,

dehydration, alkylation, andolefination converted 124 tohirsutene (Scheme 44).[50b] Thistandem approach to hirsutenecan be modified to an asym-metric version when the aldolreaction used a chiral cata-lyst.[50b] The cycloadduct 123can also be transformed to un-symmetric diketone 125, which

then can be further elaborated to the synthesis of pentale-nene and asterisca-3(15)-diene (Scheme 44).[50c]

Barluenga[51] reported the first Dotz-like reaction betweenconjugated dienyl carbene chromium complexes and termi-nal alkynes to afford eight-membered carbocycles(Scheme 45). This reaction can be named as a [(5+1)+2] re-action since the five-carbon synthon and CO come from127, and the external alkyne acts as a two-carbon synthon.

Scheme 39.

Scheme 40.

Scheme 41.

Scheme 42.

Scheme 43.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1085

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

This reaction provides an elegant way to synthesize 4/8 bicy-clic ring compounds, which are challenging to all above-mentioned metal-catalyzed/mediated cycloadditions. Bar-luenga found that this [(5+1)+2] reaction can also be per-formed photochemically with similar reaction yields com-pared to those obtained under thermal reaction conditions.

6. ACHTUNGTRENNUNG[5+3] Cycloadditions

Liebeskind and co-workers developed h3-pyranyl and h3-pyr-idinyl molybdenum p-complexes 129 as five-carbon synthonsin cycloadditions (Scheme 46).[52] In 2003, they discoveredthat, under catalysis of TMSOTf or Sc ACHTUNGTRENNUNG(OTf)3, 129 reactedwith oxyallyl cation 132, in situ generated from 130, to giveoxa- and azaACHTUNGTRENNUNG[3.3.1]bicycles 131 via a [5+3] reaction. Theyfound such a [5+3] reaction produced enantiocontrolledcomplexes 131 when enantiopure complexes 129 were used.The [5+3] reaction was proposed to occur via a stepwiseprocess starting from the attack of the in situ generatedcation 132 toward 129, giving rise to a cationic intermediate133. Then intramolecular ring-closure gives the final [5+3]products 131 with exclusive endo selectivity. Complexes 131can be transformed to various functionalized hetero-bicycliceight-membered carbocycles via demetalations (such asusing strong acids, ceric ammonium nitrate (CAN) with or

without Et3N, or pyridiniumdichromate). Liebeskind foundthat in some cases, the reactiongave [3+2]cycloadducts as by-products.

7. ACHTUNGTRENNUNG[7+1] Cycloadditions

Sarel and Langbeheim foundan Fe(CO)5 mediated [7+ 1]reaction between divinylcyclo-propanes and Fe(CO)5 underphotolysis conditions(Scheme 47).[53] This reactionwas proposed to occur via theintermediate of ferra-cyclono-nadiene 135, which, upon lossof CO and Fe(CO)3, gavecyclo-octadienones 136 and

137. This [7+1] reaction also gave organometallic com-plexes 138 and 139 as the byproducts.

Scheme 44.

Scheme 45.

Scheme 46.

Scheme 47.

1086 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.

Conclusions and Perspectives

In this Focus Review, we present the recent progresses inmetal-catalyzed cycloadditions to synthesize eight-mem-bered carbocycles, which are core structures of many naturalproducts of biological and medicinal significance. The ad-vantages of these metal-catalyzed cycloadditions are: usuallythe reactions use easily accessed starting materials; the reac-tion conditions are mild; moderate to high reaction yieldscan be achieved in most cases. These metal-catalyzed reac-tions have great impact on the synthesis of natural productswith eight-membered carbocycle skeletons or skeletonsformed from eight-membered carbocycles, as demonstratedin the synthesis of astericanolide and the skeleton of ophio-bolin F by Wender via Ni-catalyzed [(4+4)] reaction,[20] ofcedrene by Rigby via Cr-mediated [6+2] reaction,[39] of hir-sutene, desoxyhypnophiline, pentalenene, and asterisca-3(15)-diene by Yu via Rh-catalyzed [(5+2)+1] cycloaddi-tion.[50]

Despite the above successes, continuous endeavors are re-quired for further advances. One such required endeavor isto develop asymmetric versions of these cycloadditions sothat the asymmetric synthesis of eight-membered carbocylescan also be easily fulfilled. The second effort in this line isto reduce the catalyst loading so that these reactions couldbe used on a large scale in the chemical or pharmaceuticalindustry. Thirdly, more applications of these reactions to thesynthesis of complex natural products should be pursued tobetter understand the scope of these reactions, and to ad-dress challenges in natural product synthesis.

Fourthly, investigating the mechanisms of these metal-cat-alyzed cycloaddition reactions is critical, not only for under-standing these reactions and providing in-depth informationfor the substrate scope expansion and reaction optimization,but also for guiding discovery and development of new reac-tions and catalysts. Two examples of eight-membered cyclo-octenone synthesis inspired by Wender�s [(5+2)] and[(6+2)] reactions and Murakami�s [4+ACHTUNGTRENNUNG(2+2)] reaction aregiven in Scheme 48, where Shair developed a hydroacyla-tion/cyclopropane cleavage strategy, and Aissa developed ahydroacylation/cyclobutane cleavage strategy to synthesizeeight-membered carbocycles.[54]

Fifthly, it is very important to test whether these metal-catalyzed reactions can be extended to the synthesis ofeight-membered heterocycles, which are also important

motifs in natural products. Two very recent examples in thisregard are Rovis�s N-heterocyclic [4+2+2] reaction[55] andWoerpel�s heterocyclic [6+2] reaction of vinyl silaoxetanes(obtained from silylenes and vinyl epoxides) and alde-hydes.[56,57] Finally, discovering and developing more newmetal-catalyzed cycloadditions that can complement or sur-pass the above-mentioned cycloadditions for the synthesis ofeight-membered carbocycles is still highly demanded, con-sidering the various substitution patterns in natural productsand pharmaceuticals.

Acknowledgements

We are indebted to generous financial support from a new faculty grantfrom the 985 Program for Excellent Researches in Peking University, theNational Natural Science Foundation of China (20825205-National Sci-ence Fund for Distinguished Young Scholars, 20772007 and 20672005),and the National Basic Research Program of China-973 Program(2010CB833203).

[1] T. Oishi, Y. Ohtsuka, A. Rahman, Studies in Nature ProductsChemistry, Vol. 3, 1989, pp. 73.

[2] For reviews about natural products with eight-membered skeletonsand the synthetic methods of synthesis of eight-membered carbocy-cles, see: a) N. A. Petasis, M. A. Patane, Tetrahedron 1992, 48, 5757;b) M. Lautens, W. Klute, W. Tam, Chem. Rev. 1996, 96, 49; c) G.Mehta, V. Singh, Chem. Rev. 1999, 99, 881; d) L. Yet, Chem. Rev.2000, 100, 2963; e) I. Nakamura, Y. Yamamoto, Chem. Rev. 2004,104, 2127; f) A. Deiters, S. F. Martin, Chem. Rev. 2004, 104, 2199;g) M. D. Mcreynolds, J. M. Dougherty, P. R. Hanson, Chem. Rev.2004, 104, 2239.

[3] D. J. Faulkner, Nat. Prod. Rep. 1988, 5, 613.[4] D. G. I. Kingston, J. Org. Chem. 2008, 73, 3975.[5] R. S. Daum, S. Kar, P. Kirkpatrick, Nat. Rev. Drug Discovery 2007,

6, 865.[6] a) G. Illuminati, L. Mandolini, Acc. Chem. Res. 1981, 14, 95; b) C.

Galli, L. Mandolini, Eur. J. Org. Chem. 2000, 3117.[7] a) M. E. Maier, Angew. Chem. 2000, 112, 2153; Angew. Chem. Int.

Ed. 2000, 39, 2073; b) A. Michaut, J. Rodriguez, Angew. Chem.2006, 118, 5870; Angew. Chem. Int. Ed. 2006, 45, 5740; c) For an ex-ample of applying Cope rearrangement in total synthesis of naturalproduct asteriscanolide, see: J. Limanto, M. L. Snapper, J. Am.Chem. Soc. 2000, 122, 8071.

[8] W. Reppe, O. Schlichting, K. Klager, T. Toepel, Liebigs Ann. Chem.1948, 560, 1.

[9] H. tom Dieck, M. Svoboda, J. Kopf, Z. Naturforsch. 1978, 33b, 1381.[10] Evans reported in 2002 the first metal-catalyzed [(2+2)+ACHTUNGTRENNUNG(2+2)] cy-

cloaddition of diyne, albeit a single example was reported. See refer-ence [25 a].

[11] a) P. A. Wender, J. P. Christy, J. Am. Chem. Soc. 2007, 129, 13402;b) P. A. Wender, J. P. Christy, A. B. Lesser, M. T. Gieseler, Angew.Chem. 2009, 121, 7823; Angew. Chem. Int. Ed. 2009, 48, 7687; c) A.Goswami, T. Ito, N. Saino, K. Kase, C. Matsuno, S. Okamoto, Chem.Commun. 2009, 439.

[12] C. Zhuo, H. Wang, G. Zhao, Synthesis 2009, 11, 1785.[13] a) H. W. B. Reed, J. Chem. Soc. 1954, 1931; b) K. Ziegler, E. Holz-

kamp, H. Breil, H. Martin, Angew. Chem. 1955, 67, 426.[14] W. Brenner, P. Heimbach, H. Hey, E. W. Muller, G. Wilke, Liebigs

Ann. Chem. 1969, 727, 161.[15] a) A. Tenaglia, P. Brun, B. Waegell, J. Organomet. Chem. 1985, 285,

343; b) P. Brun, A. Tenaglia, B. Waegell, Tetrahedron Lett. 1983, 24,385.

[16] a) H. tom Dieck, J. Dietrich, Chem. Ber. 1984, 117, 694; b) H. tomDieck; J. Dietrich, Angew. Chem. 1985, 97, 795; Angew. Chem. Int.Scheme 48.

Chem. Asian J. 2010, 5, 1072 – 1088 � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemasianj.org 1087

Transition-Metal-Catalyzed Cycloadditions for Eight-Membered Carbocycles

Ed. Engl. 1985, 24, 781; c) M. Mallien, E. T. K. Haupt, H. tomDieck, Angew. Chem. 1988, 100, 1091; Angew. Chem. Int. Ed. Engl.1988, 27, 1062.

[17] K. Itoh, K. Masuda, T. Fukahori, K. Nakano, K. Aoki, H. Nagashi-ma, Organometallics 1994, 13, 1020.

[18] P. A. Wender, N. Ihle, J. Am. Chem. Soc. 1986, 108, 4678.[19] P. A. Wender, M. L. Snapper, Tetrahedron Lett. 1987, 28, 2221.[20] a) P. A. Wender, N. C. Ihle, C. R. D. Coeeria, J. Am. Chem. Soc.

1988, 110, 5904; b) P. A. Wender, J. M. Nuss, D. B. Smith, A. Su�rez-Sobrino, J. V�berg, D. Decosta, J. Bordner, J. Org. Chem. 1997, 62,4908.

[21] M. Murakami, K. Itami, Y. Ito, Synlett 1999, 951.[22] P. H. Lee, K. Lee, Y. Kang, J. Am. Chem. Soc. 2006, 128, 1139.[23] a) A. Greco, A. Carbonaro, G. Dall’Asta, J. Org. Chem. 1970, 35,

271; b) A. Carbonaro, F. Cambisi, G. Dall’Asta, J. Org. Chem. 1971,36, 1443; c) A. Takahashi, T. Inukai, J. Chem. Soc. Chem. Commun.1970, 1473; d) J. E. Lyons, H. K. Myers, A. Schneider, Ann. N. Y.Acad. Sci. 1980, 333, 273.

[24] a) M. Lautens, W. Tam, C. Sood, J. Org. Chem. 1993, 58, 4513; b) M.Lautens, W. Tam, J. C. Lautens, L. G. Edards, C. M. Crudden, A. C.Smith, J. Am. Chem. Soc. 1995, 117, 6863.

[25] a) P. A. Evans, J. E. Robinson, E. W. Baum, A. N. Fazal, J. Am.Chem. Soc. 2002, 124, 8782; For its diastereoselective version usingtemporary silicon tether, see: b) P. A. Evans, E. W. Baum, J. Am.Chem. Soc. 2004, 126, 11150.

[26] a) P. A. Evans, E. V. Baum, A. N. Fazal, M. Pink, Chem. Commun.2005, 63; For computational study of the reaction mechanism, see:b) M.-H. Baik, E. W. Baum, M. C. Burland, P. A. Evans, J. Am.Chem. Soc. 2005, 127, 1602.

[27] a) S. R. Gilbertson, B. Deboef, J. Am. Chem. Soc. 2002, 124, 8784;b) B. Deboef, W. R. Counts, S. R. Gilbertson, J. Org. Chem. 2007,72, 799.

[28] a) S. R. Gilbertson, G. S. Hoge, Tetrahedron Lett. 1998, 39, 2075;b) S. R. Gilbertson, G. S. Hoge, D. Genov, J. Org. Chem. 1998, 63,10077.

[29] P. A. Wender, J. P. Christy, J. Am. Chem. Soc. 2006, 128, 5354.[30] M. Murakami, S. Ashida, T. Matsuda, J. Am. Chem. Soc. 2006, 128,

2166.[31] G. Hilt, J. Janikowski, Angew. Chem. 2008, 120, 5321; Angew. Chem.

Int. Ed. 2008, 47, 5243.[32] J. A. Varela, L. Castedo, C. Saa, Org. Lett. 2003, 5, 2841.[33] R. E. Davis, T. A. Dodds, T.-H. Hseu, J. C. Wagnon, T. Devon, J.

Tancrede, J. S. McKennis, R. Pettit, J. Am. Chem. Soc. 1974, 96,7562.

[34] a) K. Mach, H. Antropiusova, P. Sedmera, V. Hanus, F. Turecek, J.Chem. Soc. Chem. Commun. 1983, 805; b) K. Mach, H. Antropiuso-va, L. Petrusova, V. Hanus, F. Turecek, P. Sedmera, Tetrahedron1984, 40, 3295.

[35] T. Schmidt, F. Bienewald, R. Goddard, J. Chem. Soc. Chem.Commun. 1994, 1857.

[36] a) D. G. Bourner, L. Brammer, M. Green, G. Moran, A. G. Orpen,C. Reeve, C. Schaverien, J. Chem. Soc. Chem. Commun. 1985, 1409;b) M. Green, S. M. Heathcock, T. W. Turney, M. P. D. Mingos, J.Chem. Soc. Dalton Trans. 1977, 204.

[37] a) S. �zkar, H. Kurz, D. Neugebauer, K. G. Kreiter, J. Organomet.Chem. 1978, 160, 115; b) C. G. Kreiter, E. Michels, H. Kurz, J. Orga-nomet. Chem. 1982, 232, 249; c) S. �zkar, K. G. Kreiter, J. Organo-met. Chem. 1985, 293, 229; d) C. G. Kreiter, Adv. Organomet. Chem.1986, 26, 297.

[38] a) J. H. Rigby, J. A. Henshilwood, J. Am. Chem. Soc. 1991, 113,5122; b) J. H. Rigby, H. S. Ateeq, N. R. Charles, J. A. Henshilwood,K. M. Short, P. M. Sugathapala, Tetrahedron 1993, 49, 5495; c) J. H.Rigby, G. Ahmed, M. D. Ferguson, Tetrahedron Lett. 1993, 34, 5397;d) J. H. Rigby, V. P. Sandanayaka, Tetrahedron Lett. 1993, 34, 935;e) J. H. Rigby, F. C. Pigge, M. D. Ferguson, Tetrahedron Lett. 1994,35, 8131; f) J. H. Rigby, P. Sugathapala, M. J. Heeg, J. Am. Chem.Soc. 1995, 117, 8851; g) J. H. Rigby, M. A. Kondratenko, C. Fiedler,Org. Lett. 2000, 2, 3917; h) J. H. Rigby, S. B. Laurent, Z. Kamal,M. J. Heeg, Org. Lett. 2008, 10, 5609.

[39] J. H. Rigby, M. Kirova-Snover, Tetrahedron Lett. 1997, 38, 8153.[40] a) M. Achard, A. Tenaglia, G. Buono, Org. Lett. 2005, 7, 2353; b) N.

Toselli, D. Martin, M. Achard, A. Tenaglia, T. Brgi, G. Buono,Adv. Synth. Catal. 2008, 350, 280; c) M. Achard, M. Mosrin, A. Te-naglia, G. Buono, J. Org. Chem. 2006, 71, 2907; d) G. Hilt, A. Paul,C. Hengst, Synthesis 2009, 3305.

[41] A. Tenaglia, S. Gaillard, Angew. Chem. 2008, 120, 2488; Angew.Chem. Int. Ed. 2008, 47, 2454.

[42] P. A. Wender, A. G. Correa, Y. Sato, R. Sun, J. Am. Chem. Soc.2000, 122, 7815.

[43] M. M. Montero-Campillo, J. Rodr�guez-Otero, E. M. Cabaleiro-Lago, Tetrahedron 2008, 64, 6215.

[44] The catalyst species was proposed as [Rh(CO)Cl] in the[Rh(CO)2Cl]2-catalyzed Wender [5+2] reactions. See: Z.-X. Yu,P. A. Wender, K. N. Houk, J. Am. Chem. Soc. 2004, 126, 9154.

[45] a) Y. Kuninobu, A. Kawata, K. Takai, J. Am. Chem. Soc. 2006, 128,11368; b) Y. Kuninobu, A. Kawata, M. Nishi, S. Yudha, J. Chen, K.Takai, Chem. Asian J. 2009, 4, 1424.

[46] P. A. Wender, G. G. Caber, R. D. Hubbard, L. Zhang, J. Am. Chem.Soc. 2002, 124, 2876.

[47] A. W. Hermann, A. de Meijere, P. A. Wender, J. Am. Chem. Soc.2005, 127, 6530.

[48] Y. Wang, J. Wang, J. C. Su, F. Huang, L. Jiao, Y. Liang, D. Z. Yang,S. W. Zhang, P. A. Wender, Z.-X. Yu, J. Am. Chem. Soc. 2007, 129,10060.

[49] F. Huang, Z.-K. Yao, Y. Wang, Y. Wang, J. Zhang, Z.-X. Yu, Chem.Asian J. 2010, DOI:10.1002/asia.201000053.

[50] For application of the [(5+2)+1] cycloaddition in total synthesis,see: a) L. Jiao, C. Yuan, Z.-X. Yu, J. Am. Chem. Soc. 2008, 130,4421; b) X. Fan, M.-X. Tang, L.-G. Zhuo, Y. Q. Tu, Z.-X. Yu, Tetra-hedron Lett. 2009, 50, 155; c) X. Fan, L.-G. Zhuo, Y. Q. Tu, Z.-X.Yu, Tetrahedron 2009, 65, 4709.

[51] J. Barluenga, F. Aznar, M. A. Palomero, Angew. Chem. 2000, 112,4514; Angew. Chem. Int. Ed. 2000, 39, 4346.

[52] R. G. Array�s, L. S. Liebeskind, J. Am. Chem. Soc. 2003, 125, 9026.[53] a) S. Sarel, M. Langbeheim, J. Chem. Soc. Chem. Comm. 1977, 827;

b) S. Sarel, Acc. Chem. Res. 1978, 11, 204.[54] a) A. D. Aloise, M. E. Layton, M. D. Shair, J. Am. Chem. Soc. 2000,

122, 12610; b) D. Crpin, J. Dawick, C. A�ssa, Angew. Chem. 2010,122, 630; Angew. Chem. Int. Ed. 2010, 49, 620.

[55] a) R. K. Friedman, T. Rovis, J. Am. Chem. Soc. 2009, 131, 10775;b) R. T. Yu, R. K. Friedman, T. Rovis, J. Am. Chem. Soc. 2009, 131,13250.

[56] M. Prevost, K. A. Woerpel, J. Am. Chem. Soc. 2009, 131, 14182.[57] For other two recent examples using 2-alkenylazetidines as four-

atom synthons in eight-membered heterocycle synthesis, see: a) S.Koya, K. Yamanoi, R. Yamasaki, I. Azumaya, H. Masu, S. Saito,Org. Lett. 2009, 11, 5438; b) B. Drouillat, F. Couty, V. Razafimaha-leo, Synlett 2009, 3182.

Received: December 14, 2009

1088 www.chemasianj.org � 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Asian J. 2010, 5, 1072 – 1088

FOCUS REVIEWSZ.-X. Yu et al.