Nitro and Other Electron Withdrawing Group Activated Ruthenium Catalysts for Olefin Metathesis Reactions

Abstract Advanced applications of the Nobel Prize winning olefin metathesis reaction require user‐friendly and highly universal catalysts. From many successful metathesis catalysts, which belong to the two distinct classes of Schrock and Grubbs‐type catalysts, the subclass of chelating‐benzylidene ruthenium complexes (so‐called Hoveyda–Grubbs catalysts) additionally activated by electron‐withdrawing groups (EWGs) provides a highly tunable platform. In the Review, the origin of the EWG‐activation concept and selected applications of the resulting catalysts in target‐oriented synthesis, medicinal chemistry, as well as in the preparation of fine‐chemicals and in materials chemistry is discussed. Based on the examples, some suggestions for end‐users regarding minimization of catalyst loading, selectivity control, and general optimization of the olefin metathesis reaction are provided.

their synthetic versatility.O lefin metathesis has been the subject of numerous general and specialized reviews, [12][13][14][15][16][17] and the reader is asked to refer to the recently published books on metathesis,a st hey can provide aw ell-organized view on this field. [1][2][3] 2. The EWG-Activation Concept, Early Observations, and Consequences Historically,t he first Hoveyda-Grubbs catalysts [18] that were substituted in the 2-(alkoxy)benzylidene ligand with groups such as -OC(O)R and -Br were made in an attempt to immobilize the parent catalysts Ru8 and Ru9. [19,20] Although the idea of using these functional groups (being just anchors for immobilization) to control the initiation rate of the homogeneous catalysts was not considered at that time,t he key observation was disclosed in 2002 when the strongly electron-withdrawing nitro group was installed para to the chelating isopropoxy fragment of Ru9. [21,22] Unexpectedly,t his small structural alteration led to al arge change in the activity:t he nitro-activated Ru12 (Figure 2) was found to be visibly more active than the parent catalyst, initiated at 0 8 8C, and gave very good results in an umber of challenging metathesis reactions. [21,22] This observation was found to be true for many other electron-withdrawing substituents,such as -SC 4 F 9 ,-SO 2 Ar, -C(O)R and -P(O)R 2 , [23][24][25] as well as -SO 2 -(Ru17,F igure 3), placed in the para or meta position relative to the chelating oxygen atom. [24,26] In fact, the same effect was observed with quaternary ammonium groups, [27] or even protonated amines and the in situ generated carbocations [28]    We reasoned that the role of the electron-withdrawing group (EWG) is to decrease the electron density on the oxygen atom of the chelating iPrO fragment, thus weakening the strength of the RuÀOb ond ( Figure 2, insert). [21] This makes the corresponding catalysts activate faster and prevent them from entering into an inactive "sleeping" state through aso-called "boomerang" mechanism. [18,29] In an independent study,Blechert and co-workers reported on aset of Hoveyda-Grubbs catalysts substituted with electron-donating groups (EDGs,mostly ethers;deactivating) and EWGs (F,CF 3 ,CN; activating) and interpreted differences in their activities with the aid of s Hammett constants (Table 1). [30] Butenschçn and co-workers disclosed an interesting and highly active bimetallic catalyst (Ru13), where aC r(CO) 3 fragment in the parene complex was used to induce both steric and electronic (EWG) activation of the parent Hoveyda-Grubbs catalyst ( Figure 2). [31] Based on the above-described EWG-activating effect, more catalysts have been logically developed and eventually commercialized, including those bearing SO 2 NR 2 , [32] NHCOR [33] (Figure 3), and PO-(OR)Ar [34] EWGs.
Then itro catalyst Ru12 [36] and Mauduitsa ctivated catalysts, [37] such as Ru15 and Ru16,h ave been reviewed previously and, therefore, in the present Review only new facts and the most instructive examples of their use will be provided. To the best our best knowledge,t he third most popular EWG analogue (EWG = SO 2 NR 2 ), the Zhan-1B catalyst Ru14,i si ncluded in areview for the first time.

Synthesis of Advanced Biologically Active and
Natural Compounds by Utilizing EWG-Activated Catalysts

Examples in Ring-Closing Metathesis
Total synthesis sometimes allows aconclusion to be made that ar eported structure of ag iven natural product requires revision. [38][39][40] This was the case in the recent disclosure from Chan and Koide on the first total synthesis of the reported structure of the heat shock protein expression inhibitor Stresgenin B( Scheme 2). [41] Thes ynthesis features an umber of synthetically challenging transformations,and among them the one-pot [42] ring-closing metathesis/oxidation event using ap roperly chosen, not-too-high amount of nitro catalyst Ru12,a nd then MnO 2 as the oxidant ensured ah igh yield (84 %) of the required intermediate.This is anice example of aw ell-planned and skillfully executed but rather straightforward ring-closure metathesis (RCM) reaction.
Theg roup from Keio University reported an improved total synthesis of incednam (2), the aglycon of the 24membered macrolactam glycoside antibiotic Incednine. [43] Theretrosynthetic analysis of 2 was based on the construction of the 24-membered macrocycle by ac hallenging intramolecular macrocyclization of the fragile polyunsaturated substrate 1 (Scheme 3). Conditions for the RCM were rigorously explored, such as screening of anumber of catalysts,including  first-and second-generation Grubbs and Hoveyda-Grubbs catalysts,a sw ell as nitro catalyst Ru12.T hese tests revealed that the best conditions consist of using Ru12 in the presence of p-methoxyphenol (PMP) [44] and 3 molecular sieves. Despite such measures,I ncednam (2)w as obtained in only 17 %o verall yield after cleavage of the TES groups (Scheme 3). Then ext example shows another challenging RCM reaction, where the laborious optimization of the substrate structure,a sw ell as aw ise selection of the catalysts and conditions changed the initial failure into asuccess.Researchers from the University of California in San Francisco reported the total synthesis of Virginiamycin M2, am ember of the streptogramin natural product group. [45] Theretrosynthesis featured al ate-stage macrocyclization to form the 23-membered ring of Virginiamycin (Scheme 4). Different cyclization reactions were tried, and after the failure of aP d-catalyzed Stille coupling, the authors turned to Ru-catalyzed olefin metathesis.T he initial efforts were rather disappointing,a st he reactions of bis-terminal precursor 3a with first-and second-generation Grubbs and Hoveyda-Grubbs catalysts at 23 8 8Cresulted in no conversion, whereas ah igher temperature (70 8 8C) resulted in the substrate being consumed but no cyclic product detected (probably because of the presence of the rather fragile conjugated diene fragment). Thus, the authors resorted to alteration of the substrate structure.T he transmethyl-substituted precursor 3b still led to unsatisfactory results with Grubbs catalysts,b ut the expected macrocycle was finally observed in the reaction mixture when the more robust second-generation Hoveyda-Grubbs catalyst (Ru9)was used. Although the yield of only 15 %h ad little practical utility,t his result showed that the RCM macrocyclization is indeed possible,a nd just required more detailed optimization. Therefore, the authors switched to the cisisomer 3c and tested aw ider set of metathesis catalysts.A lthough most of them (such as the polymerization catalysts) did not improve the yield of the target macrocycle,i tw as found that ab atch-wise addition of acatalyst (2 8instead of 20 mol %added in one portion) and am ore polar solvent (PhCF 3 instead of PhCH 3 )l ed to small improvements. [46,47] Under these conditions,the best catalysts were EWG-activated Ru14 and Ru12,w hich provided the expected macrocycle in yields of 28 and 49 %, respectively.
Interestingly,when the desilylated substrate (3d)was used in the RCM reaction instead of 3c,t he productivity of the key macrocyclization reaction increased greatly,with Virginiamycin M2 (4)isolated in 72 %yield when using Ru12 in CH 2 Cl 2 at room temperature (Scheme 4). [45] Although the cause of this improvement is unclear, the authors suggested that it can arise from the coordination of the catalyst to the exposed allylic alcohol or from favorable conformational bias in the "unprotected" macrocyclic precursor 3d relative to 3c. [45] Whatever the reason for this fortuitous effect is,t he reported study shows the importance of ac areful optimization of the metathesis step,b oth by adjustments made in the substrate structure and by the right choice of the catalyst. This fact can be seen in numerous examples in the literature, [3] and the proficient cooperation of organic synthetic chemists with experts in metathesis appears to be the easiest way to optimize such challenging metathesis processes.
Another example where the key RCM step is ar eal challenge was reported by Christmann and co-workers in the total synthesis of the RNAp olymerase inhibitor Ripostatin B. [48] Thea uthors planned to construct the sensitive 14membered macrolactone,which features the peculiar doubly skipped triene,b ym eans of RCM. Unfortunately,anumber of problems were encountered with popular catalysts,such as first-and second-generation Grubbs or Hoveyda-Grubbs Ru9,i ncluding loss of the E/Z selectivity,s luggish reactions, or olefin truncation (to form 7;Scheme 5). Thelast "parasitic process", caused by cross-metathesis with ethylene produced during the reaction, strongly depended on the nature of the catalyst. Whereas the Ru2 catalyst (20 mol %) led to two products (both of them useless for obtaining the target molecule)-E/Z-6 and truncated 7-in an approximately 1:1 ratio,t he Dorta catalyst [49] Ru21 (10 mol %) showed ap erverse selectivity,giving 7 as the major product (but as asingle E isomer!).C areful analysis of these failures led to an improvement. Using catalyst Ru12 together with tetrafluoro-1,4-benzoquinone (8)k nown for its anti-isomerization properties, [50] and purging with argon minimized the problems encountered previously and allowed the expected product 6 to be obtained in an acceptable yield and with full Eselectivity in the CÀCd ouble-bond formation (Scheme 5). [48] Interestingly,t he authors used an isocyanide reagent [51] to quench the Ru metathesis catalyst just after the RCM was completed, thus preventing unwanted metathesis side reactions. [52][53][54] Wiese and Hiersemann reported another challenging RCM macrocyclization in the synthesis of natural and nonnatural Jatropha-5,12-dienes. [55] Ther egioselective RCM reaction of triene 9 was expected to establish the fully substituted 12-membered trans-bicyclo[10.3.0]pentadecane framework of the target compounds.U nfortunately,i nt he initial trials it was found that the first-generation Grubbs catalyst led to no conversion and the second-generation Hoveyda-Grubbs catalyst delivered alow (22 %) yield of the expected 3-epi-characiol 10.T othe authors relief,the crucial RCM could then be realized using catalysts Ru2 and Ru12, which afforded, after removal of the remaining protecting groups,t he expected key product 10,w hich was later transformed into anumber of natural and non-natural members of the Jatrophane family of diterpenes (Scheme 6). [55,56] Them etathesis of substrates containing basic nitrogen atoms (e.g. alkaloids precursors) are known to be difficult and require the use of acid additives to prevent coordination of the amine to the ruthenium center, which can result in catalyst poisoning.R esearchers from the Va nderbilt Center for Neuroscience Drug Discovery reported av ery interesting total synthesis of the Stemona alkaloid Stemaphylline,its C9aepimer,and their N-oxides. [57] Ambitiously,itwas planned to close both the 5-and 7-membered rings of these natural products in two independent one-pot RCM events (Scheme 7). Unfortunately,t he conditions (Ru2,t rifluoroacetic acid) developed for other azabicyclic ring systems [58] proved to be unsuccessful in the case of tetraene 11. Fortunately, Ru12 in the presence of the same acid (1 equiv) gave the desired product 12,albeit in low yield. Interestingly, this challenging RCM reaction was found to be sensitive to the nature and amount of the acid used to protonate the basic nitrogen atom present in 11.T he best results were achieved with two equivalents of camphorsulfonic acid (CSA). Using optimized conditions for closure of the azepine ring, the authors tried the far more ambitious tandem bis-ring closure to obtain bis-RCM product 13 in one reaction. As mall selection of metathesis catalysts were screened again (Schrock molybdenum catalyst and an umber of Ru catalysts: Ru2, Ru9, Ru12, Ru14). From this set, complexes Ru9 and Ru12 were found to be optimal, and finally the reaction of tetraene 11 in the presence of CSA was conducted with Ru12 to give 13 in 52 %y ield (from 11). This product was then converted in high yield into 9a-epi-Stemaphylline (Scheme 7) and 9a-epi-Stemaphylline N-oxide. [57] Encouraged by this success,t he researchers approached the synthesis of Stemaphylline,analkaloid which differs from 9a-epi-Stemaphylline in the configuration at just one stereocenter. [59,60] Surprisingly,this relatively small change made the analogous RCM reaction fail. Thea uthors put al ot of effort into understanding the reasons for such alarge difference in reactivity between the 9a-epi-Stemaphylline and Stemaphylline tetraene precursors,a nd finally opted to use the relay ring-closing-metathesis (RRCM) strategy to save the project. [61] RRCM is an important technique that was developed to force some recalcitrant substrates to enter the metathesis cycle or to differentiate between competitive metathesis pathways. [61,62] Typically,i tc an be carried out by introducing as pecial relay-arm into as ubstrate molecule,t ow hich the Rucarbene moiety can attach more easily,a nd then undergoing as equence of intramolecular transformations (including the release of as table cyclic by-product) to yield, at the end, the desired cyclic olefin (see insert in Scheme 7). Luckily,t he application of the RRCM method to the spe- cially designed substrate 14 led (after additional optimization steps) to formation of the desired Stemaphylline precursor 15, although in moderate yield (37 %; Scheme 7, bottom). Hiersemann and co-workers reported an elegantly designed total synthesis,w here RCM was used to close both the 5-and the 14-membered rings of (À)-9,10-Dihydroecklonialactone B. [63] Thee nvisioned formation of the 14-membered lactone seemed rather straightforward (RCM of ar ather simple bis-terminal diene);u nfortunately,t his endeavor turned out to be much more challenging than expected. After screening an umber of catalysts and conditions,t he nitro catalyst Ru12 (5 mol %) in the presence of 1,4-benzoquinone (17;0.1 equiv) at elevated temperature was found to perform best (Scheme 8). To avoid decomposition, this rather unstable intermediate was immediately subjected to Zn-mediated b-elimination to deliver the lactone 18 in 41 %yield (after two steps). This was transformed into triene 19 for the second RCM event. Thes econd RCM (this time using catalyst Ru2 to promote the cyclopentene formation) provided the bicyclic core of the target molecule.F urther standard transformations provided access to (À)-9,10-Dihydroecklonialactone B(Scheme 8).
Thee xamples outlined above,a nd those that had to be skipped because of limited space, [64][65][66][67][68][69][70][71] testify the importance of the detailed optimization of aR CM reaction, but also shows that it is sometimes rather difficult to rationalize why two seemingly very similar catalysts or substrates give substantially different results. [72]

Examples in Cross-Metathesis Reactions
Cross-metathesis reactions were for al ong time considered as technically more difficult than ring-closing metathesis. [73] With the introduction of modern olefin metathesis catalysts,c rossmetathesis reactions with a,b-unsaturated compounds,s uch as acrylic esters,a crolein, or vinyl ketones became possible and now are frequently used in target-oriented syntheses.The advantage of this approach is that the two reacting partners exhibit different character and reactivity.I np articular, the less-reactive electron-deficient olefin typically does not enter into parasitic self-metathesis reactions (Scheme 1), and can be used in excess,t hus allowing for high selectivity and ah igh yield with cross-metathesis.F rom the numerous Ru catalysts available,E WG-activated complexes such as Ru12 or Ru14 usually give good results. [74][75][76][77] Less-active substituted a,b-unsaturated compounds,s uch as methacrolein, can also be utilized in cross-metathesis. Researchers from the University of Pittsburgh reported the synthesis of unnatural Bistramide Aa nalogues,w ith the purpose of comparing their potencyw ith that of the natural product. [78] TheR CM step consisted of ar eaction between spirocycle 20 and 550 equivalents methacrolein in the presence of nitro catalyst Ru12,which led to the product 21 in an acceptable yield (Scheme 9). This universal intermediate was then used to obtain an umber of Bistramide Aa nalogues. [78] Fürstner and co-workers reported the concise synthesis of the putative structure of the highly cytotoxic marine macrolide Mandelalide A. [79,80] Cross-metathesis between terminal alkene 22 and functionalized enone 23 worked very well in the presence of Ru14 and furnished the required enone building block 24 with high E-selectivity (Scheme 10).
Similarly,t he successful cross-metathesis between advanced enone building block 25 and functionalized alkene 26 mediated by nitro-activated Ru12 was reported in the synthesis of the cytotoxic spiroketal Spirangien A( Scheme 11). [81] Thecase where two reacting olefins are of similar reactivity is more complicated, as each partner can independently undergo "dimerization" through homometathesis (Scheme 1), thereby limiting the yield of the key cross-metathesis reaction, and giving am ixture of products that is often complicated to separate.T he following examples show how to deal with such ap roblem.
Thes tructurally unique FR901 464 was isolated at the Fujisawa Pharmaceutical Company from the culture broth of

Angewandte Chemie
Reviews ab acterium of Pseudomonas species,a nd proven to possess antitumor activity against an umber of cell lines,t hus having the potential for clinical application. [82] Koide and co-workers targeted the synthesis of FR901464 and their analogues. [83] After the failure of an umber of synthetic strategies,i nt he final attempt, Koide and co-workers envisaged forming the key CÀ Cdouble bond by cross-metathesis as the very last step in the synthesis. [83] As shown in Scheme 12, cross-metathesis between 27 and 28 was not at rivial task, and the proper choice of the catalyst was the key.T he fragile nature of 28 prevented using more forcing reaction conditions,b ecause the reacting partners quickly decompose above 47 8 8C. Therefore,t he conditions identified by Koide and co-workers involved the use of 12 mol % Ru12 in 1,2-dichloroethane (DCE) at precisely 40 8 8C. Thec ross-metathesis partner 28 was used in an excess of 1.8 equivalents to maximize the consumption of the other partner (27). This allowed FR901464 to be obtained in 40 %y ield after one recycle of the unreacted starting materials (51 %y ield based on recovered 27). Encouraged by the success of the above strategy,K oide and co-workers decided to synthesize the FR901464 analogue Meayamycin, which was achieved in 59 %y ield by the same strategy after one recycle of the recovered starting materials (Scheme 12). [83] Later, more than adozen other analogues of FR901464 were prepared by using similar Ru12-catalyzed late-stage cross-metathesis reactions, and some of them have been shown to be significantly more potent than Meayamycin against several cancer cell lines and, therefore,ofinterest in oncology. [84] Nicolaou et al. reported the efficient and selective total syntheses of natural products exhibiting arelated structure-Thailanstatins A-C. [85] En route to these advanced targets,the full orchestra of transition-metal-catalyzed transformations-including an umber of olefin metathesis events-was used but, unlike in Koides retrosynthesis,t he key step to combine the two advanced fragments of the natural product was aS uzuki coupling reaction, not metathesis.T he boronate precursor (32)f or the key Suzuki step was,h owever, made by a Ru12-catalyzed crossmetathesis with vinyl boronate 31 (Scheme 13). This successful cross-metathesis of the rather challenging boronate partner 31 was then repeated multiple times in the synthesis of numerous analogues of Thailanstatin, thereby allowing for evaluation of their cytotoxicity against anumber of cancer cell lines. [85] Cross-metathesis with vinyl borolanes is ap opular maneuver in target-oriented syntheses. [73] Another example is the conversion of the chiral homoallylic alcohol (PMB-protected) 33 into trans-vinylborolane 34 in high yield and stereoselectivity (E/Z > 20:1; Scheme 14). [86] Anumber of cross-metathesis events between vinyl boronates and advanced olefinic building blocks cata- lyzed by Ru2 and Ru12 have also been used in the synthesis of Rhizopodin. [87] Another impressive example where two complex alkene fragments are combined to form atarget molecule in the last step has been presented by Hahn and co-workers in the total synthesis of Projerangolid and Jerangolid E. [88] Reacting an excess of the less-reactive olefin 35 with diene 36 in the presence of the second-generation Grubbs catalyst led to the isolation of Jerangolid Ei n2 3% yield. In contrast, using Ru12 in perfluorinated toluene [46,89,90] delivered the target compound in 93 %y ield and excellent E-selectivity (Scheme 15). Thee stablished cross-metathesis conditions were then applied for the synthesis of Projerangolid, as well as for non-natural 5-epi-Projerangolid and 9-(Z)-Jerangolid E. [88] When the direct cross-metathesis reaction is for some reasons impossible or not selective enough, RCM of at emporary tether can lead to better results.Prunet and co-workers tried this strategy recently. [91] During the previous studies toward the synthesis of Dolabelide C, the C16-C30 fragment (Scheme 16) of this natural product was obtained using ac rossmetathesis reaction. [92] Unfortunately,d espite al arge amount (45 mol %) of the Hoveyda-Grubbs catalyst having been used, the cross-metathesis reaction yield was only 47 %, thus showing the limitations of cross-metathesis for the synthesis of highly hindered trisubstituted olefins.I nt he more recent approach, Prunet and co-workers used asilicon-tether RCM strategy to obtain the same C16-C30 fragment of Dolabelide C. An interesting catalyst influence was noted in this RCM reaction (Scheme 16). Whereas second-generation Grubbs catalysts afforded an already satisfactory yield of 63 %(compared with the previous cross-metathesis-based synthesis), the Hoveyda-Grubbs catalyst improved it to 76 %. TheZ han-1B catalyst was the least productive,w hile the nitro and the Mauduit catalysts were superior, with an impressive 81 %y ield of 39 obtained with the Umicore M71 SIMes complex. [91] There are more examples of cross-metathesis that are worth discussing, [74][75][76][77][93][94][95][96] but the limited space of this Review does not allow this.

Examples in Enyne Cycloisomerization
One of the many nice examples where enyne cycloisomerization was utilized in the synthesis of advanced natural products is Hondasl ong-term study on Securinega alkaloids. [97] These naturally occurring alkaloids exhibit attractive biological activities and constitute an ambitious synthetic target. In one of the retrosynthetic approaches, Honda et al. intended to employ at andem one-pot enyne cycloisomerization followed by ar ing-closing metathesis (enyne-RCM) of enyne acrylic ester 40 as the key reaction (Scheme 17, top). However,t he treatment of this substrate with Ru12 afforded the d-lactonic compound (42)i nstead of the desired compound 41.C learly,aruthenium carbene generated from the terminal alkene in the butenyl group reacted with the alkyne prior to the C À Cdouble bond of the a,b-unsaturated ester ( 2 -then-1 instead of 1 -then-2 ,s ee Scheme 17 top). To solve the problem, al ess-reactive alkene moiety (bearing as acrificial ethyl end group) and amore reactive allyl ether (instead of the acrylic ester) were introduced in the modified substrate 43.I nt his case,t he tandem enyne-RCM reaction proceeded very well with only 2mol % of the highly active catalyst Ru12 and gave the properly cyclized product (44)i n7 4% yield (Scheme 17, bottom). Thus,t he stereoselective construction of the remaining rings was achieved in ar elatively short sequence,c ompleting the first synthesis of enantiomerically pure (À)-Securinine. [98] A similar sequence was then used to obtain (+ +)-Viroallosecurinine. [99] Thea bove enyne cycloisomerization followed by RCM sequence was later modified in av ery clever way by Yang,L i, and co-workers to obtain other Securinega alkaloids:( À)-Norsecurinine,( + +)-Allonorsecurinine,( À)-Flueggine A, and (+ +)-Virosaine B. [100] Thek ey enyne cycloisomerization followed by RCM reactions were tested using several commercially available metathesis catalysts,i ncluding second-generation Grubbs and Hoveyda-Grubbs catalysts, but the EWG-activated Zhan-1B catalyst (Ru14)p rovided the best results. [100]

Applications in the Synthesis of Active Pharmaceutic Ingredients
Applications of olefin metathesis in the preparation of active pharmaceutic ingredients (APIs) have recently been reviewed. [17,[101][102][103] Therefore,o nly af ew of examples will be described herewith, as they nicely show current challenges related to the use of metathesis technology in the pharmaceutical industry.
Thehepatitis Cvirus (HCV) is amajor cause of chronic liver disease and can lead to cirrhosis,c arcinoma, and liver failure. TheW orld Health Organization (WHO) estimates that 130-170 million people are chronically infected with the HCV,a nd is al eading cause of liver transplants. [104] A number of detailed studies describing the design, structure-activity relationship studies,s cale-up synthesis,a nd clinical trials of novel macrocyclic HCV protease inhibitors have recently been published. Interestingly, despite av ariety of macrocylization methods having been tried, one of the most frequently used-also in large-scale production-was the RCM reaction Angewandte Chemie Reviews ( Figure 4). Ciluprevir,t he first such macrocycle reported, [105] failed in clinical trials;n evertheless,i ts synthesis remains important as the first commercially viable large-scale RCM macrocyclization that influenced an umber of subsequent synthetic approaches to other anti-HCV macrocycles.T he original scale-up reaction leading to the formation of the Ciluprevir cyclic precursor was carried out at high dilution with 5-7 mol %ofthecatalyst Ru8.Key to amore economical process was the modification of the substrate structure by installation of aB oc protecting group on the proline NH amide fragment (marked in bold in Figure 4a). This seemingly small change significantly reduced the ring strain, and allowed RCM at concentrations 10-20 times higher than those used previously.Aswitch to an EWG-activated catalyst further improved the process economy,a so nly 0.1 mol % Ru12 was needed to obtain a93% yield of 45. [106] In summary, the researchers at Boehringer Ingelheim put al ot of effort into optimizating the RCM process,n ot only focusing on practical issues but, importantly,u nderstanding its reaction mechanism and kinetics. [107] Thus,s ynthetic strategies optimized for Ciluprevir,s uch as controlling the effective molarity with N-protecting groups,w ork not only in closely related cases,such as the formation of the macrocyclic core of BI201302, [108] but were also useful in RCM leading to less related APIs,s uch as Simeprevir [109] and others. Researchers from GlaxoSmithKline and Anacor Pharmaceuticals Inc. reported the serendipitous discovery of anovel and potent HCV protease inhibitor as ab y-product in the synthesis of another antiviral molecule. [110] Like some of the predecessors (e.g.D anoprevir), the macrocyclic urea derivative 46 was obtained in ahigh yield by the RCM reaction, this time in the presence of the Zhan-1B catalyst under highdilution conditions (c = 0.01m,F igure 4b)a nd with ah igh catalyst loading (20 mol %). [110] Then ext example is the preparation of IDX316 (Figure 4c). [111] Instead of using the previously utilized strategy,i nt his case the metathesis reaction was implemented as the final step to generate the API. Theevident advantage of late-stage RCM is the shorter and less complicated synthetic route that has al ower cost contribution of the RCM step to the overall cost of the synthesis.T he potential disadvantage is that removal of the toxic [112] Ru catalyst at the API step can be more difficult and lead to contamination of the product with metal traces, especially as the RCM reaction was conducted with ar elatively large loading of 1.25 mol %ofthe Zhan catalyst Ru14. After comprehensive studies on the removal of the catalyst on ak ilogram scale,acombination of treatment with triphenylphosphine oxide and ad imercaptotriazine solid-supported scavenger was utilized to reduce the Ru level to < 10 ppm. [111] Merck recently reported the multi-kilogram synthesis of the MK6325 drug candidate,i nw hich the RCM step was also made at the late stage of the synthesis,w hen the heteroaromatic part, typical for this class of molecules,h ad already been installed (Figure 4d). [113] Although MK6325 consists of two macrocyclic fragments,o nly one of them was formed by RCM. Them etathesis-based macrocyclization to afford 48 was effected in toluene at rather high dilution (0.06 m)bythe slow addition of Zhan-1B (0.45 mol %) at 80 8 8C, whereby the presence of benzoquinone was mandatory to inhibit isomerization of the CÀCdouble bond. [113] Not only has RCM been utilized in pharmaceutical R&D and scale-up laboratories.L uesch and co-workers disclosed at the beginning of 2008 the structure of Largazole,anovel peptide-polyketite hybrid. [114] It was isolated in trace amounts from amarine cyanobacteria of the genus Symploca (reclassified now as the new genus Caldora penicillata)c ollected at Key Largo, Florida. Largazole displays very potent growth inhibition activity in several transformed human and murine-derived cell lines.M any research groups targeted the synthesis of Largazole,a nd these advances have been reviewed. [115] Among them, the groups of Hong, [116] Cramer, [117] and Phillips [118] ambitiously opted to introduce the complete thioester side chain by ac ross-metathesis reaction in the very last step of the synthesis (Scheme 18). Such an approach has ab uilt-in advantage of avoiding the need for several protection group manipulations,a nd an additional strategic benefit consists of the simplified generation of analogues (needed for SAR studies) just by exchange of the cross-metathesis olefinic partner (e.g. 50). However,asfar as the original structure of Largazole is concerned, one must consider the risks related to cross-metathesis with an olefin bearing as ulfur substituent in the chelating position (Scheme 18, insert). This constitutional difficulty was reflected in the high catalyst loading (20-50 mol %) used in all published syntheses and the modest yields obtained with second-generation Grubbs and Hoveyda-Grubbs catalysts. Thebest solution found by Cramer and co-workers was to use the more active nitro catalyst Ru12,which allowed Largazole to be obtained in an acceptable 75 %yield. Recently,Oceanyx Pharmaceuticals Inc. reported the development of the scaledup synthesis of Largazole by using Cramersconditions for the crucial cross-metathesis step and utilizing Ru12 (Scheme 18). [119] In the developed process,d ecagrams of Largazole were synthesized in an overall yield of 21 %for the longest linear sequence.
Researchers at Boehringer Ingelheim Pharmaceuticals and Astatech BioPharmaceutical Corporation, supported by experts on olefin metathesis from Apeiron Synthesis S.A., reported the stereoselective synthesis of substituted 1,4benzodioxanes that are structural motifs of several drugs. [120] As the pharmaceutical industry has to keep the chemical production process as competitive and as cost-effective as possible,alot of effort was invested in optimizing each of the individual reaction steps.RCM of substrates containing vinyl ethers is known to be problematic, and ah igh loading of the second-generation Grubbs catalyst (5-8 mol %) was previously reported for the synthesis of 1,4-benzodioxines, [121] which hindered the practicality of this transformation. Gratifyingly,amounts of nitro catalyst Ru12 as low as 150 to 300 ppm were found to lead to the target 1,4-benzodioxanes in > 80 %y ield (Scheme 19).

Production of Specialty Chemicals and Commodities, as well as Applications in Materials Science
To the best of our knowledge,t he nitro-and sulfonamide-activated(Zhan-1B) catalysts of the EWG-activated family of catalysts are only commercially available as their first-generation versions (L = PCy 3 ). Although the Hoveyda-Grubbs catalysts with SIMes or SIPr ligands are undoubtedly more stable and active, [18,33] there is alimited number of potential applications for the first generation of EWG-activated catalysts.One of them is enyne metathesis of acertain class of substrates bearing an internal acetylenic bond. [22,122] It was observed in this case that the use of second-generation Grubbs or Hoveyda-Grubbs catalysts usually led to the formation of undesired products (e.g. 53; Scheme 20). Asimilar lack of selectivity has been reported by Mori and co-workers for the second-generation Grubbs catalyst bearing an IMes ligand. [123] Interestingly,t he firstgeneration nitro catalyst (Ru22)s hows ah igh level of selectivity in this transformation, leading only to the formation of 52.T his observation is also true in the case of other enynes possessing an internal alkyne motif. [22,122] EWG-activated catalysts have also been tried in ADMET polymerization (Scheme 1) for the synthesis of poly(p-phenylenevinylenes) (PPVs)-important materials for applications in organic light-emitting diodes (OLEDs) and organic photoconductors.I nt his study,s elected divinylbenzene and divinylfluorene monomers were polymerized under vacuum using Ru2 or Ru12 olefin metathesis catalysts to afford PPVs as free-standing films (Scheme 21). [124] Then itro-activated

Angewandte Chemie
Reviews 13750 www.angewandte.org catalyst has also been used in the synthesis of selectively substituted indenes,w hich have potential applications in photovoltaics. [125] An important increase in the scope of applications of the EWG-activated catalysts was achieved through further modifications of the catalysts structure.T he iodide-containing nitro catalysts were synthesized by scientists from the company Apeiron Synthesis S.A. and applied in an umber of challenging RCM and cross-metathesis reactions,a lso in "green" solvents (Scheme 22). [126] It was noted that the augmented steric hindrance in the vicinity of the Ru center (because of the higher ionic radius of the iodide ligands compared with chloride ligands) ensures the higher stability and robustness of the catalyst. This benefit was best illustrated under highly challenging conditions,such as in reactions with very low catalyst loading, in protic (MeOH, iPrOH) or in Lewis-basic solvents (2-MeTHF), or in the presence of various impurities or ethylene.T he presence of ethylene (instantaneous removal of which is sometimes difficult to achieve under industrial large-scale conditions) sometimes has ad ramatic effect on the reaction yield and selectivity,a s illustrated in Scheme 22. Thei ncreased stability of the ruthenium methylidenes generated from Ru23 makes this catalyst especially suitable for the macrocyclization of unbiased dienes,s uch as 54 with low effective molarity,a lso in the presence of ethylene. [126] Another potentially important direction in the further improvement of EWG-activated catalysts lies in modification of the NHC ligand. Although both SIMes (initiating faster) and SIPr (more stable) [33,127] versions of the leading EWG catalysts (M71, M73, and nitro) are commercially available,o ther NHC modifications are less obvious,b ut some of them are very interesting and will be described here.B uchmeiser and co-workers reported the synthesis of nitro catalyst Ru27 bearing a1 ,3-bis(2,4,6trimethylphenyl)-3,4,5,6-tetrahydropyrimidin-2-ylidene ligand ( Figure 5). This catalyst was tested in anumber of RCM, cross-metathesis,and ROMP transformations,and was also immobilized on asolid support, which led to avery low Ru contamination of the products. [128] Apeiron Synthesis S.A. developed au nique class of selfscavenging olefin metathesis catalysts bearing apolar quaternary ammonium group installed on the backbone of standard SIMes or SIPr ligands. [16] Thep resence of this group allows efficient separation of ruthenium impurities after the reaction. Ar epresentative member of this class is complex Ru28 [129] (Figure 5), which in addition is water soluble and can also promote olefin metathesis in aqueous media. [14,130] Thea pplication of catalyst Ru28 led to products that exhibited low ruthenium contamination levels after as imple and inexpensive purification step,c onsisting only of water extraction or filtration through as hort pad of silica. [129] Thea ctive Ru species generated from ruthenium complexes bearing standard SIMes or even SIPr NHC ligands exhibit limited stability under certain demanding conditions (high dilution, high temperature,presence of ethylene). [103] As ac onsequence,i ndustrial implementation of modern expensive metathesis catalysts in al arge-scale production of commodity materials (where the low product price excludes technically complicated and cost-intensive approaches) is problematic.R CM production of macrocyclic musks,s elective homometathesis of a-olefins,o re thenolysis of biosourced fatty oils are examples of processes were the great potential of metathesis is hampered by the low stability of the available catalysts.Over the past 19 years,alarge number of ruthenium complexes bearing NHC ligands have been obtained, but most of them cannot be used at ppm levels in the abovementioned processes because of their insufficient lifetime under demanding reaction conditions in the presence of various impurities.B ased on excellent results reported by Bertrand, Grubbs, and co-workers on cyclic alkyl amino carbene (CAAC) ligands, [131] Skowerski and co-workers disclosed the EWG-activated ruthenium CAAC complex Ru29,which promoted the difficult RCM macrocyclization leading to am usk-smelling lactone at ac atalyst loading of 30 ppm, and crossmetathesis of acrylonitrile at 25 ppm (Scheme 23). [132] Thel atter result, in particular, which was obtained in cooperation with the company Arkema, is of interest, as acrylonitrile

Angewandte Chemie
Reviews was for years considered to be avery difficult cross-metathesis partner and reactions with this partner usually required the use of industrially unacceptably high amounts of metal. [133][134][135] In an in-depth study,Nascimento and Fogg explained the fundamental reasons for the remarkable productivity of Ru29,s tressing the fact that while the CAACc atalysts are more resistant to a-elimination, they are very susceptible to bimolecular decomposition-a well-known path of destruction for many Ru catalysts.Importantly,however, because the CAAC catalysts can be used at very low catalyst loading, the bimolecular decomposition is inhibited under these conditions,t hereby making them extraordinarily productive. [136] Thet ranslation of ar eaction from the laboratory to process scale using traditional batch techniques is sometimes very challenging.I nt he specific case of olefin metathesis, as trong dependence of the reaction yield on the reactor design and scale was noted many times. [103] Exploring this area, researchers from Snapdragon Chemistry Inc. and Massachusetts Institute of Te chnology developed as pecial continuous-flow reactor featuring am embrane sheet-inframe pervaporation module that enables effective removal of ethylene.Under these conditions,the diiodo complex Ru23 and UltraNitroCat Ru29 gave particularly good results in RCM macrocyclizations relevant to the fragrance industry. [137] Recently,Dorta and co-workers reported the preparation of nitro catalysts bearing sterically augmented NHC ligands. [138] Importantly,c omplex Ru30 demonstrated quite good activity in the formation of tetrasubstituted C À Cdouble bonds,t he reaction which was traditionally the Achilles heel of nitro catalyst Ru12 (Scheme 24). [21,22] We hope that the collected examples illustrate the potential of the application of EWGactivated Hoveyda-Grubbs catalysts in the synthesis of various building blocks and fine-chemicals,a sw ell as in materials science.

Practical Considerations, Outlook, and Perspectives
Thes pecialized EWG-activated catalysts,t ogether with the already very successful classical Schrock and Grubbs-type catalysts,constitute apowerful toolkit that allows synthetic chemists to perform even very challenging metathesis transformations.H owever, the variety of metathesis catalysts now commercially available makes the proper choice of the catalyst at rue problem. Thed ata reported herein demonstrate that great care must be taken when choosing the appropriate catalyst for ag iven metathesis reaction. Thea vid reader certainly has identified an umber of examples where more than one metathesis catalyst type has been used in ag iven total synthesis (e.g.o ne catalyst for cross-metathesis and the other one later for RCM;s ee Section 3.1). [57,63,85,86] Interestingly,a nd in contrast to ag rowing number of applications of EWG-activated catalysts in total synthesis and medicinal chemistry,t hese complexes have found only limited use in polymer production through ROMP,w here other catalyst types dominate. [139][140][141][142][143] We can only repeat what we stated already in 2008:different metathesis catalysts prove to be optimal for different applications and no single catalyst can outperform all others in all cases. [72] Therefore,d uring optimization of especially important (or industrial) metathesis processes,i ti ss uggested to screen all major types of catalysts available,o no neso wn or with the help of metathesis experts.O ur long-term experience advises that more close cooperation between synthetic chemists ("end-users") and the catalyst developers or manufactures can substantially speed-up and smooth the costly optimization phase,especially in the case of complicated industrial projects.F or example, the typical loading of ametathesis catalyst used in academiapublished total syntheses of natural products is from 5t o 25 mol %o re ven higher (up to 100 mol %). [144] This is,o f course,f ully understandable in academic research, and even justified at the early stages of the industrial R&D work, but sooner or later the loading must be substantially reduced to make the production process economically viable.S ome examples where the amount of catalyst used was reduced from amultiple molar percent to its decimal fractions or even to ppm levels was achieved, thanks to the cooperation with the catalyst producer or metathesis expert, have been presented in this Review. [98,106,120] In this context, recent results by Nascimento and Fogg are quoted again, as they were able to prove that lowering the loading of the CAAC nitro catalyst is not only favorable from an economic point of view,b ut in fact it also increases the catalystss tability,b y inhibiting the bimolecular decomposition pathway. [136] This result shows yet again the importance of fundamental studies to deepen our understanding of the factors that influence the stability and activity of catalysts.I nt he practice of total synthesis leading to complex polyfunctional natural products and in medicinal chemistry,the prediction of the most optimal catalyst for ag iven substrate is not an otiose question but rather as erious problem. We hope that ab etter understanding of the initiation and decomposition mechanisms of catalysts will help to solve this riddle.
Thei mportance of the proper choice of as olvent (both classical petroleum-based solvents and more eco-friendly "green" ones are available nowadays), [145] the beneficial influence of various additives (e.g.p henols and quinones), [50] the surprising effect of fluorinated aromatic solvents, [46,47,[89][90] and many other "enabling techniques" [146] have also been identified. Importantly,i tisnot only through the use of such sophisticated techniques,b ut even changing the simplest reaction parameters,s uch as temperature and concentration, switching to abatch-wise addition of the catalyst, or just more efficient removal of ethylene,c an sometimes significantly improve the outcome of the metathesis reaction.
In as ummary,w eh ave tried to convince the reader that EWG-activated Ru catalysts have enabled, and will continue to enable,s yntheses of various chemical molecules in many fields of organic and medicinal chemistry.