Synthesis of (−)‐Dihydroraputindole D by Enantioselective Benzoylation of a 1,3‐Diol Intermediate

Abstract The enantioselective synthesis of (−)‐dihydroraputindole D is reported. The key step is the desymmetrizing benzoylation of a prochiral 1,3‐diol employing Trost′s ProPhenol catalyst system, which has been applied for the first time to a cyclic molecule carrying geminal hydroxymethyl groups. The cyclopenta[f]indoline system was assembled by Au(I)‐catalyzed cyclization of an alkynylated indoline precursor. (−)‐Dihydroraputindole D was obtained in 17 steps and 8% overall yield starting from dihydroxyacetone. In combination with quantum chemical calculations of the ECD spectra, our synthesis allowed us to determine the absolute configuration (5S,7R) of the natural product (+)‐raputindole D from the Rutaceous plant Raputia simulans.

In the course of our total synthesis of raputindole A( 1), we developed the regioselective Au I -catalyzed cyclization of 6-alkynylindole precursors as the key step in constructingt he cyclopenta[f]indole system. [10] Having the quaternary centeri n-stalled,d iastereoselectivity of the synthesis was achievedb y tethered Ir-catalyzed hydrogenationo fa ni ntermediatec yclopentene-containing tricycle. [11] However, our attempts to assemble the quaternary center in an enantioselective manner by using chiral Au I complexes remained unsuccessful.
For raputindole D( 2), optical activity was reported without determination of the absolute configuration. Raputindole Ddiffers from raputindole Abythe presence of ahydroxymethyli nstead of am ethyl group at the quaternary center (C5, Scheme 1). Thisl ed us to the idea of exploiting the 1,3-diol moiety of cyclopenta[f]indoline 5 for an enantioselectived esymmetrization strategy.A mong the few existing methods, the use of Trost's ProPhenol catalyst 4 appeared to be most promising, as the benzoylation of 2-monosubstituted1 ,3-diols had provided yields and enantioselectivities superiort ot hose accessiblebye nzymatic methods. [12] In the case of indole 5,i tw as unclear how catalyst 4 would behaveb oth in terms of reactivity and enantioselectivity.I n the mechanism proposed by Trost et al.,the enantiodifferentiation of the two hydroxymethyl groups is possible because the carbonyloxygen of the benzoyl source-vinyl benzoate-coordinatest ot he zinc center from the sterically clearly preferred side, where the hydrogen substituent is located (R' = H, Scheme 1). [12b] However,d iffering froma ll earlier examples, our envisaged substrate 5 contains aq uaternary carbon in the 2positionoft he 1,3-diol moiety.
To our surprise, the assignment of the absolute configuration of tricyclic benzoate 15 by quantum chemical calculation of the ECD (TDDFT, wB97XD/TApr-cc-pVDZ)p roved to be difficult. Fortunately,t he situation was clearf or the tricyclica lkenyl triflate 13,f or which the configuration shown in Scheme 3w as unambiguously assigned by ECD calculation (see the Supporting Information). The prediction of the stereochemical outcome of the benzoylation was not possible based on analogy with existing examples.C ompounds 13 and 14 constitute the first examples with aquaternary centerint he 2-position.
The successfulm onobenzoylation of 11 gave us the opportunity to exploit the remaining primary hydroxy group as a tetherf or the diastereoselectiveC rabtreeh ydrogenation of tricyclic diene 15.T oo ur surprise it proved to be necessaryt o use the high amount of 28 mol %o fC rabtree catalyst [Ir(COD)py(PCy 3 )]BARF [15] to achieve as atisfactory degree of conversion (Table 1). It was impossible to achieve the selective monohydrogenation of 15.E ven at À20 8C, the major product was the isobutyl-substituted indane derivative 17 (79 %), which was ac-companied with isobutenyl indane 16.A tÀ40 8C, we observed only minimal conversion, whereas at 0 8C, dihydrogenation of had taken place exclusively.
When starting from diol 14,t he hydrogenation was much faster,b ut it wass till impossible to avoid reduction of the isobutenyl side chain. Theb est result was obtained when using [Ir(COD)py(PCy 3 )]PF 6 ,p roviding a1 :2 mixture of mono-and dihydrogenated products, accompanied by traces of startingm aterial. Chiral hydrogenationc atalysts (entries 6-8, Ta ble 2) were not successful, either.
To our surprise, hydrogenation of 14 in the presenceo f [Ir(COD)(S)-tBu-PHOX)]BARF (5 mol %) on the 10 mg scale afforded the diastereomeric Diels-Alder dimers 20 a (59 %) and 20 b (7 %) as racemic major products(Scheme 4), the structures of which were elucidated by extensive 2D NMR spectroscopy. In both cases, we observed as inglet for the aliphatic methine-Ho ft he cyclohexenem oiety (d = 3.15, 3.14 ppm), whiche xcludes the alternative regiochemistry.T he decisive NOESYc orrelationa llowing the assignment of the major diastereomer was observed between6 -H and 8'-H. By DFT calculation (B3LYP/6-31G(d)), diastereomer 20 a is more stable than 20 b. For the formation of 20 a and 20 b,h alf of the starting material must have undergone isomerization of the isobutenylcyclopentene to a2 -methylallylidene moiety. One of the rare examples, where this behavior occurred when employing aCrabtree catalyst under hydrogenation conditions, was described by Guillou et al. [16] who reported the isomerization of an exocyclic methenyl double bond to the endocyclic position. The hydrogen served only as activator of the Crabtreec atalyst, but was not incorporated.
Given the difficulties experienced with the monohydrogenation of 15,wed ecided to pursue the enantioselective synthesis of dihydroraputindole D( 3), which differs from the natural product by the presence of an isobutyl insteado fa ni sobutenyl side chain.
In summary,w ereport the first enantioselective synthesis of ar aputindole derivative. Key steps are the Au I -catalyzed cyclization forming the cyclopenta[f]indoline system and the enantioselective benzoylation of the achiral tricyclic 1,3-diol 11 employing Trost's ProPhenol catalyst system.T hus, in addition the synthesis dihydroraputindole D, our approach also addresses a hitherto unexplored typeo fs ubstrate for ProPhenol-type catalysts. Desymmetrization of the 1,3-diolp rovedt ob es uperior to hydrogenation employing ac hiral Crabtreec atalyst that led to isomerization and surprisingd imerization of the 1,3-diol precursor.( À)-Dihydroraputindole D( 3)w as obtainedi n1 7s teps and 8% overall yield starting from dihydroxyacetone. Our synthesis also allowed us to determine the absolutec onfiguration of the natural product (+ +)-raputindole D( 2)f rom the Rutaceous plant R. simulans.

Note added in proof
After revisiono ft his manuscript, an ew total synthesis of (+)-raputindole Awas reported. [17]