Converting a birch reduction product into a polyketide: Application to the synthesis of a C1-C11 building block of rimocidin

FULL PAPER

DOI: 10.1002/ejoc.201201112

Converting a Birch Reduction Product into a Polyketide: Application to the

Synthesis of a C¹–C¹¹Building Block of Rimocidin

Luc Nachbauer^[a]and Reinhard Brückner*^[a]

Dedicated to Professor Herbert Mayr on the occasion of his 65th birthday

Keywords: Asymmetric synthesis / Reduction / Ozonolysis / Epoxides / Spiro compounds / Organocopper compounds

A stereoselective synthesis of a C¹–C¹¹building block for the

polyol-polyene antibiotic rimocidin has been developed. Its

functional groups originate from a disubstituted indane,

which underwent Birch reduction, and oxidative cleavage.

This provided a dihydropyranone with a β-keto ester side-

chain. The latter was subjected to a Noyori hydrogenation

(ds Ͼ 97:3). An oxy-Michael addition gave a mixture of two

spiroketals. Luche reduction then led to three spiroketals in

an 80:10:10 ratio. The major spiroketal became isolable by

separating one of the by-products chromatographically and

the other by a diastereomer-selective thioketalization. The

remaining spiroketal was ring-opened to give a dithiolane.

Its CO₂Me group was converted into the 1,3-dithiane unit of

the target compound (i.e., 47) using an odor-reducing work-

up procedure, which should prove to be generally useful.

Introduction

Rimocidin (1a) is a polyol-polyene macrolide that was

isolated from Streptomyces rimosus in 1951.^[1]The presence

of a through-conjugated tetraene moiety and an aminogly-

coside – derived from d-mycosamine – were recognized by

Cope et al.^[2](Scheme 1). Work by the groups of Rinehart^[3]

and Borowski^[4]culminated in a proof of the complete

structural formula of rimocidin (1a) in 1995. A tetrahydro-

pyrancarboxylic acid moiety (C¹²–C¹⁸fragment) is a note-

worthy substructure of rimocidin because it is also present

in several other polyol-polyene macrolides,^[5]namely in am-

photericin B,^[6]nystatin A₁,^[7]candidin,^[8]and pimar-

icin.^[9,10]Each of these compounds shows antifungal ac-

tivity.^[11]Nonetheless – and in contrast to amphotericin B,

nystatin A₁, and pimaricin – rimocidin (1a) is used neither

in human therapy^[12]nor in food preservation.^[13]

Total syntheses of the aglycon (i.e., 1b) of rimocidin (1a)

have been achieved in the laboratories of Rychnovsky^[14a]

and Smith III.^[14b]Similar successes have been reported

concerning amphotericin B,^[15]nystatin A₁,^[16]candidin,^[17]

and pimaricin.^[18]Our group has studied several variations

of the preparation of the common tetrahydropyrancarbox-

ylic acid moiety (C¹²–C¹⁸fragment) mentioned above.^[19,20]

In this paper, we report how this activity has been extended

[a] Institut für Organische Chemie und Biochemie,

Albert-Ludwigs-Universität Freiburg,

Albertstraße 21, 79104 Freiburg, Germany

Fax: +49-761-2036100

E-mail: reinhard.brueckner@organik.chemie.uni-freiburg.de

Homepage: http://ww w .brueckner.uni-freiburg.de/

Supporting information for this article is available on the

WWW under http://dx.doi.org/10.1002/ejoc.201201112.

Scheme 1. Retrosynthetic analysis of C¹–C¹¹synthon 2 of the po-

lyol-polyene antibiotic rimocidin (1a) and its aglycon 1b.

6904

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

to the stereoselective synthesis of the equivalent (i.e., 47; from a C_Ar–OMe motif in the starting material. It is note-

Scheme 9) of a C¹–C¹¹synthon 2 (Scheme 1) for rimocidin worthy that alcohol, silyl ether, ketal, and carboxylic acid

(1a).

moieties survive the combined reaction conditions of

Scheme 2.

Retrosynthetic Analysis

Macrolides are usually obtained by the Yamaguchi or a

related macrolactonization of an appropriately protected

seco-hydroxy acid.^[21]As a rule, such macrolactonizations

rely on a convergent approach, in which a number of build-

ing blocks are first assembled independently and joined

later. Envisaging a synthesis of rimocidin (1a) in that man-

ner, we identified three appealing precursors (Scheme 1):

(1) a C²⁰–C³⁰or C¹⁹–C³⁰fragment (not shown) for estab-

lishing the tetraene scaffold through a cross-coupling or an

olefination; (2) a C¹²–C¹⁹(not shown) or C¹²–C¹⁸counter-

part that represents the “eastern” moiety (e.g., the epoxy

alcohol 3, which is reported elsewhere^[19a]); and (3) a syn-

thetic equivalent of C¹–C¹¹synthon 2. The latter was con-

ceived as an aldehyde derivative with umpolung reactivity

that would be able to attack an epoxide (e.g., 3) at C¹²nu-

cleophilically. Thus, we planned to install a 1,3-dithiane in-

stead of the aldehyde function^[22]in the synthetic equivalent

46 of synthon 2 (cf. Scheme 9).

It was planned that the hydroxylated stereocenters of C¹–

C¹¹synthon 2 would come from the regio- and stereoselec-

tive bis-reduction of hydroxy-triketo ester 4 (Scheme 1).

The C=O bonds of 4 were envisaged to be the result of a

double oxidative cleavage of the alkene and enol ether moie-

ties of substituted cyclohexa-1,4-diene 5. This compound

looked like the product of a Birch reduction^[23]of disubsti-

tuted indane 8. Its stereocenter was deemed to be accessible

by stereoselective alkylation of arylmetal 6 by enantiomer-

ically pure triflate 7, or of an ethylmetal by enantiomerically

pure epoxide 9.

Scheme 2. Previously obtained^[26–29]polyfunctional building blocks

for polyketide synthesis from sequences starting with a Birch re-

duction, which delivers an annulated cyclohexa-1,4-diene, and end-

ing by an ozonolysis. Heteroatom-free non-oxygenated aromatics

lead to polycarbonyl compounds, while mono- and dimethoxy-sub-

stituted aromatics contribute one ester group or two, respectively.

Our Birch reduction/ozonolysis approach is unusual.

This is because it uses both entities that result from the C=C

bond cleavage. In ozonolyses of “standard” Birch reduction

products, the cyclohexa-1,4-dienes do not have an annu-

The literature precedent shown in Scheme 2 encouraged

lated ring. Such an ozonolysis completes the conversion of our belief that the Birch reduction/ozonolysis sequence

an aromatic precursor into a 1,3-diketone or a β-keto es- 8Ǟ4 that we planned (Scheme 1) stood a fair chance of

ter.^[24]In contrast, a Birch reduction product like 5 success – even though 4 was a triketo ester, and no such

(Scheme 1), which contains an annulated ring, gives a teth- species had been obtained by this methodology prior to our

ered ozonolysis product. Such a product contains twice the work. On the other hand, the precedent shown in Scheme 2

number of C=O groups compared to an untethered ozon- alerted us to the fact that the sensitivity of polycarbonyl

olysis product from a non-annulated Birch product.^[25]This compounds akin to 11,^[26]14,^[27]16,^[28]and 20^[29]– i.e., in

suggests that the synthesis of polyoxygenated molecules our case 4 – towards decomposition, particularly in the

might become advantageous by such an approach. Despite presence of acid or base, was not to be underestimated.

this, it has not been used extensively. Pertinent examples of

In addition we were conscious of a difficulty that would

which we are aware are compiled in Scheme 2. The sub- have to be faced once we were in possession of triketo ester

strates are three indanes (i.e., 10a,b^[26]and 12^[27]), one tetra- 4: the need to differentiate the carbonyl groups at C³, C⁵,

lin (i.e., 15^[28]), and a cyclopentane-annulated dialin (i.e., and C⁹. The C³=O and C⁹=O bonds would have to be re-

19).^[29]These precursors yield Birch reduction/ozonolysis duced by hydride additions from the top face (with respect

products with four new carbonyl groups (10a,bǞ11, to the orientation of formula 4 in Scheme 1) without affect-

12Ǟ14), with two new carbonyl groups and two new ester ing the C⁵=O bond. We were confident that the C¹¹=O

functionalities (15Ǟ16), or with five new carbonyl groups bond, which is part of an ester, would not compete for the

and one new ester functionality (19Ǟ20). An ester results reductants. Finally, we were aware of the possibility of par-

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6905

L. Nachbauer, R. Brückner

FULL PAPER

ticipation by a neighbouring functional group at the stage previously been used for acylations, including a chloroacet-

of the anticipated Birch reduction/ozonolysis product 4. Its ylation, of aromatics with at least two hydroxy or methoxy

OH group looked to be poised to add to the C⁵=O bond groups.^[37]We adopted the Houben–Hoesch method to our

and form a hemiketal. When we reduced first the C⁹=O and monomethoxylated aromatic 22 by using chloroacetonitrile

then the C³=O bond, we were in fact able to exploit this (6 equiv.) as the solvent and heating the mixture at 85 °C.

hemiketal formation to protect the C⁵=O bond.

The resulting iminium chloride was hydrolyzed at reflux

temperature after adding concentrated aqueous HCl. Chlo-

roacetophenone 24 was isolated in 60% yield after a single

recrystallization from n-hexane.

Synthesis of the Enantiomerically Pure Indane

A modified^[38]Corey^[39]–Itsuno^[40]reduction^[41]of ketone

24 gave chlorohydrin 25 (Scheme 3). Compound 25 was iso-

lated in 76% yield by extractive work-up and recrystalli-

zation from n-hexane. It had an ee of 99.6%.^[42]Treatment

of this chlorohydrin with a suspension of K₂CO₃in MeOH

(room temp., 2 h) induced ring-closure. The crude product

(≈ 92% yield) consisted predominantly of the desired epox-

ide (i.e., 9). However, it contained up to 20% of a by-prod-

uct, which we interpreted as resulting from an epoxide-ring

opening by MeO^–. As epoxide 9 decomposed on silica

gel,^[35]removal of the by-product was impossible. In con-

trast, treatment of chlorohydrin 25 with aqueous NaOH in

iPrOH (room temp., 1.5 h) gave epoxide 9 cleanly in 96%

yield.

Scheme 3 shows how we bridged the gap between 5-hy-

droxyindane (21), which is commercially available, and io-

dine-substituted indane 23, which we considered as a pre-

cursor of metalated indane 6 (see Scheme 1). Hydroxy-

indane 21 was methylated with Me₂SO₄in the presence of

aqueous KOH. The resulting methoxyindane (i.e., 22; 91%

yield) reacted with iodine monochloride in AcOH^[30]to give

an 83:17 mixture of monoiodo derivatives in favor of the

ortho-substitution product 23. This ratio was improved to

95:5 by recrystallization.

Our first attempt towards indane-substituted alcohol 8

followed the 6 + 7 strategy outlined in Scheme 1, and is

shown in detail in Scheme 4. It was based on the recently

described^[43]stereoselective substitutions of the triflate

group of enantiomerically pure α-(trifluoromethanesulf-

onyl)carboxylic esters 26 upon treatment with an alkylmag-

nesium chloride and a catalytic amount of ZnCl₂(Ǟ27;

Scheme 4, top). We tried an analogous substitution reaction

between arylmagnesium chloride 6a, catalytic ZnCl₂, and

racemic α-(trifluoromethanesulfonyl)carboxylic ester 7a,

which had not been studied in ref.^[43]Arylmagnesium com-

Scheme 3. Synthesis of indanes 23 and 9 for the 6 + 7 and the 9 +

EtM approaches to the indane core, as shown in Scheme 1. Rea-

gents and conditions: a) Me₂SO₄(1.2 equiv.), KOH (50% aq.;

4.0 equiv.), acetone, reflux, 1 h; 91%. b) ICl (1.3 equiv.), AcOH,

room temp., 2 h; 49%, dr = 95:5 after recrystallization. c) Chloro-

acetonitrile (6.0 equiv.), ZnCl₂(2.5 equiv.), HCl_g, 85 °C, 2 h; HCl

(concd. aq.), reflux, 30 min; 60%. d) BH₃·THF (1.0 equiv.), (R)-

5-(hydroxydiphenylmethyl)pyrrolidin-2-one (2 mol-%), THF, room

temp., 5 h; 76%. e) NaOH (aq.; 7.0 equiv.), iPrOH, room temp.,

1.5 h; 96%.

Scheme 3 also shows how we converted the same com-

mercially available indane (i.e., 21) into indane-substituted

epoxide 9, which we needed for the preparation of indane-

substituted alcohol 8 (see Scheme 1). To this end, we sub-

jected methoxyindane 22 to a regioselective ortho-(chloro-

acetylation) to give chloroacetophenone 24. Attempted

Friedel–Crafts acylations with chloroacetyl chloride and

AlCl₃(1.1 equiv.) at various temperatures (0 °C, 20 °C,

40 °C) in CH₂Cl₂, CS₂or nitrobenzene resulted mostly in

demethylation.^[31,32]Generating the chloroacetylating agent

in situ from chloroacetic acid and either P₂O₅/MeSO₃H^[33]

[34]

or P₂O₅/SiO₂

allowed access to acetophenone derivative

Scheme 4. Testing the 6 + 7 approach to the indane core, as shown

in Scheme 1. Reagents and conditions: a) Ref.^[43]6, THF, 0 °C; ad-

dition of AlkMgCl (1.4 equiv.), ZnCl₂(5 mol-%), 2.5 h; Ͼ 70%.

b) 23 (1.4 equiv. relative to 7), iPrMgCl (1.4 equiv.), THF, –25 °C,

1.5 h; 86% conversion into 6a.^[44]c) Either 7a (racemic, 1.0 equiv.),

ZnCl₂(10 mol-%), THF, 0 °C; addition of 6a, 4 h, then room temp.,

16 h; or 7b (ee Ͼ 99.5%, 1.0 equiv.), ZnCl₂(15 mol-%), THF, 0 °C;

24 without removing the methyl group. However, by-prod-

ucts formed that were hard to remove by flash chromatog-

raphy on silica gel^[35]and subsequent recrystallizations from

n-hexane. The issue was solved with a Houben–Hoesch

acylation^[36], i.e., using chloroacetonitrile, ZnCl₂, and HCl

gas as the source of the electrophile. This procedure had addition of 6a, 2.5 h (cf. footnote^[46]).

6906 Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

Synthesis of a C¹–C¹¹Building Block of Rimocidin

pound 6a was prepared in 86% yield^[44]by an iodine/mag- trol, the desired α-attack should be favored by benzylic acti-

nesium exchange reaction^[45]between iPrMgCl and iodoin- vation, but an undesired β-attack might compete because

dane 23. However, several combinations of triflate 7a and of a lower steric hindrance. Regarding stereocontrol, first

6a did not give the desired substitution product (i.e., 28), principles suggested that observing 100% inversion of con-

but led instead to the decomposition of the triflate. We figuration would be more likely than 100% retention. For

wondered whether Grignard reagent 6a was too sterically this reason, we had synthesized R-configured styrene oxide

hindered (because 6a is α-branched, whereas other Grig- 9, not its enantiomer. Table 1 provides an overview of the

nard reagents that were used to form substitution products regioselectivity of styrene oxide reactions from the literature

27^[43]were not), not nucleophilic enough (because its carb- using organometallic reagents not too different from

anionic center is sp²- rather than sp³-hybridized), or stabi- “EtM”. Table 1 also compiles observations concerning their

lized by the methoxy group (through electron withdrawal stereoselectivity. Last but not least, Table 1 covers a chemo-

from the carbanionic center and/or by complexation of the selectivity issue that arises when a a styrene oxide 29 isom-

Mg atom). As a compensatory measure, we tried to couple erizes into an arylacetaldehyde by a semipinacol rearrange-

arylmagnesium reagent 6a with triflate 7b (Scheme 4, bot- ment,^[47,48]and the latter compound reacts with the organo-

tom). The latter compound contains a CO₂nBu group. metallic reagent to give a non-benzylic secondary alcohol

Hence, it should be less sterically hindered than the pre- 32. Grignard reagents attack styrene oxides 29 selectively at

viously used triflate (i.e., 7a), which contained a CO₂tBu C^αto give primary alcohols 30 (Table 1, entries 1, 2).^[49,50]

group. The new reactant combination failed to give the sub- However, they cause some epoxide rearrangement (Ǟ non-

stitution product 28. However, the diminished steric hin- benzylic secondary alcohols 32).^[49,50]This side-reaction

drance of triflate 7b resulted in an entirely different reac- was circumvented by using Et₂Mg (Table 1, entry 3),^[49]

tion, albeit one that was useless for our purposes.^[46]The plausibly because the Lewis acidity of the metal was attenu-

same undesired reaction occurred when we omitted the ated. Me₃Al (Table 1, entries 4–6) gave no rearrangement/

ZnCl₂or replaced it by LiCl·CuCl (10 mol-%).

addition product 32.^[51,52]Et₃Al gave a little (Table 1, en-

Our second attempt towards indane-substituted alcohol try 7), but not if PPh₃was present (Table 1, entry 8).^[53]Li-

8 followed the 9 + “EtM” strategy outlined in Scheme 1. AlBu₄did not give 32 either (Table 1, entry 9).^[54]Each of

The aim was to open styrene oxide 9 by an ethyl nucleophile these reagents opened styrene oxides 29 selectively at C^α.

regioselectively and stereoselectively. Regarding regiocon- Surprisingly, in the cases studied (Table 1, entries 4, 5),

Table 1. Ring-opening reactions of various styrene oxides 29 with organometallics collected from the literature.

Yield ee of

of 30 30

Entry

Ref.

Subst

ee

“RM”

Solvent

Yield of 31 Yield of 32

[49]

[50]

1

2

H

EtMgBr

MeMgI

Et₂Mg

Et₂O

38%

–

38%

not re-

ported

–

25%

2%

–

not re-

ported

–

2,5-OMe-4-Me

51%

[49]

[51]

[52]

[53]

[54]

[55]

[56]

[55]

[57]

[58]

[59]

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

H

Et₂O

75%

100% Me₃Al

n-hexane

85% 85%^[a]

2,3,4,5-F

2-OAr-4-I

97%

93%

Me₃Al

Et₃Al

Et₃Al, 5 mol-% PPh₃

LiAlBu₄

BuCu

Bu₂CuLi

BuCu(CN)Li

Bu₂Cu(CN)Li₂

Me₂Cu(CN)Li₂

Hex₂CuMgBr

MeCu(CN)Li,

BF₃·Et₂O

n-hexane/CH₂Cl₂(2:1) 94% 63%^[a]

hexanes

toluene

Et₂O

THF

Et₂O

75% 89%^[a]15%

H

50%

93%

92%

11%

32%

28%

74%

29%

8%

–

19%

48%

4%

21%

40%

85%

13%

60%

65%

33%

–

THF

not reported

98%

96%

67% 96%^[b]33%

[62]

[52]

19

H

THF

75% 96%^[b]

–

MeCu(CN)Li,

BF₃·Et₂O

20

2-OAr-4-I

93%

THF

70% 91%^[b]

–

[a] The major enantiomer resulted from substitution with retention of configuration. [b] The major enantiomer resulted from substitution

with inversion of configuration.

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6907

L. Nachbauer, R. Brückner

FULL PAPER

Me₃Al gave the product with 60–85% retention of configu- and with a variety of ethylmetal reagents (Table 2, entries 2–

ration.^[51,52]This was rationalized by a two-step S_Ni-mecha- 13). Vinylmagnesium bromide caused somewhat less semipi-

nism.^[51]Organocopper reagents ring-open styrene oxides nacol rearrangement (Table 2, entry 1; Ǟ36) than did ethyl-

29 with little preference for C^αor for C^βattack (Table 1, magnesium iodide (Table 2, entry 2; Ǟ37), while diethyl-

entries 9–18).^[55–59]This observation is valid for alkylcopper magnesium caused none (Table 2, entry 3). The major reac-

reagants, Gilman cuprates, Normant cuprates,^[60]“lower-or- tion mode of these reagents was to open epoxide 9 with

der cyanocuprates”, and “higher-order cyano cuprates”.^[61]perfect α- vs. β-selectivity. Nonetheless, they were useless

The highest proportions of C^αattack were seen with for our purposes, because they eroded ca. 60–80% of the

BuCu(CN)Li (Table 1, entry 13) and Me₂Cu(CN)Li₂stereointegrity of substrate 9.^[63]They substituted preferen-

(Table 1, entry 16) in diethyl ether rather than THF solu- tially with retention of configuration. So, too, did AlEt₃

tions. However, full regioselectivity could be reached with and LiAlEt₄, but again with a considerable loss of stereoin-

MeCu(CN)Li in THF in the presence of BF₃·Et₂O (Table 1, tegrity (Table 2, entries 4, 5). Moreover, epoxide rearrange-

entries 19, 20), and in these reactions, complete inversion ment^[47]was more prominent than with the Grignard rea-

(95–100%) of configuration was observed.^[52,62]An advan- gents (Table 2, entries 1, 2), and it became worse when we

tage of the organocopper reagents was that none of them used EtAlO/(Et₂Al)₂O^[64]as an ethyl donor (Table 2, en-

induced a semipinacol rearrangement that could have led try 6). Overall these observations can be explained – and

to the formation of alcohol 32.

put in perspective in the context of the results of Table 1 –

Table 2 describes our efforts to open indane-substituted by considering the Lewis acidities of Mg^IIand Al^III, and

epoxide 9 with vinylmagnesium bromide (Table 2, entry 1) the +M effect of the methoxy group in the indane-moiety of

Table 2. Ring-opening reactions of indane-substituted epoxide 9 with C₂H_n-transferring organometallics.

Yield

36 of

or 33

37 or 8

Entry

“RM”

Reaction conditions

(possibly including the preparation of “RM”)

33

or

8

34

or

35

ee^[a]of

33^[b]

or 8^[c]

VinylMgBr (2.0 equiv.), Et₂O/THF (2:1), 0 °C,

15 min; H₂(1 atm.), Pd(OH)₂/C, EtOAc

EtMgI (2.0 equiv., 1.5 m in Et₂O), Et₂O, 0 °C to

40 °C, 1 h

MgEt₂(1.1 equiv.), Et₂O, 0 °C to room temp.,

15 min

AlEt₃(2.0 equiv.), PPh₃(5 mol-%), toluene,

0 °C, 15 min

AlEt₃(1.5 equiv.), EtLi (1.5 equiv.), n-hexane,

0 °C, 15 min

1

2

3

4

5

VinMgBr

EtMgI

93

89

:

0

:

7

68%

11 55%

70%

–15%^[d]

–39%^[d]

–29%^[e]

–31%^[d]

–25%^[d]

Et₂Mg^[49]

100

66

0

[53]

AlEt₃, cat. PPh₃

34 43%

22 61%

[54]

LiAlEt₄

78

AlEt₃(10 equiv.), H₂O (6.0 equiv.), CH₂Cl₂,

–20 °C, 30 min

not deter-

mined

6

7

8

EtAlO/(Et₂Al)₂O^[64]

EtLi^[65]

49

51 27%

12%

EtLi (2.0 equiv.), Et₂O, 0 °C to room temp., 1 h

EtLi (2.0 equiv.), CuI (2.0 equiv.), Et₂O, –40 °C

to –20 °C, 30 min, 0 °C, 3 h, room temp., 16 h

EtLi (3.0 equiv.), CuI (1.5 equiv.), Et₂O, –40 °C

to –20 °C, 30 min, 0 °C to room temp., 30 min

EtLi (3.0 equiv.), CuCN (1.5 equiv.), Et₂O,

–40 °C, 4 h, room temp., 16 h

EtLi (1.2 equiv.), CuCN (1.2 equiv.), THF,

–40 °C to –20 °C, 2 h, room temp., 18 h

EtLi (1.5 equiv.), CuCN (1.5 equiv.), BF₃·Et₂O

(1.5 equiv.), THF, –78 °C, 15 min

complex mixture

+5%^[d]

EtCu

19%

0%^[e]

9

Et₂CuLi

38

77

:

62

23

:

0

24%

56%

+91%^[d]

+89%^[d]

[61]

10

11

12

13^[f]

Et₂Cu(CN)Li₂

no conversion; epoxide 9 was reisolated in quantita-

tive yield

EtCu(CN)Li^[56]

EtCu(CN)Li/

80

:

0

:

20 30%

76%

+3%^[d]

BF₃·Et₂O^[56]

EtLi (1.7 equiv.), CuCN (1.7 equiv.), Et₂O,

–40 °C to 0 °C, 4 h, room temp., 16 h

EtCu(CN)Li^[56]

100

0

+95%^[d]

[a] In this column, each ee value is preceded by the sign of the optical rotation of a CDCl₃solution (20 °C) of the isolated sample of 8

(it must be noted that the ring-opening products 33 and 36 were hydrogenated to give 8 and 37 prior to isolating 8). Levorotatory samples

of 8 resulted from substitutions with more retention than inversion of configuration, dextrorotatory samples of 8 from substitutions with

inversion of configuration. [b] We determined the preponderant absolute configuration of ring-opening product 33 after hydrogenation

of the olefinic C=C bond (Ǟ8); see footnote [c]. [c] We established that the preponderant absolute configuration in the dextrorotatory

sample of ring-opening product 8, which emerged from the experiment from entry 13 was (S) (cf. Scheme 5). [d] Determined by chiral

HPLC (details: ref.^[63]). [e] Determined by measuring [α]²_D⁰. [f] The epoxide 9 used in this experiment had 99.6% ee.

6908

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

epoxide 9. These reagents favor the formation of a benzylic α-attack was perfect. The stereochemical outcome was tan-

cation, which persists until it picks up an ethyl residue from tamount to 97.7% inversion and 2.3% retention, as con-

a face almost at random, or until it rearranges.

cluded from isolating alcohol (+)-8 (76% yield) as an (S)-

EtLi, which is probably less Lewis acidic than Mg- or enantiomer with 95.1% ee.

Al-based organometallics, and which is known to open ali-

The absolute configuration of ring-opening product (+)-

phatic terminal epoxides nucleophilically,^[65]did not not re- 8 was established as S by X-ray crystallography^[66,67]of bro-

act cleanly with indane-substituted epoxide 9. A mixture of mine-containing^[68]ester (S)-39 (Scheme 5). Compound 39

many compounds resulted, from which we isolated just 12% was prepared from (+)-8 in three steps and 24% overall

of the desired alcohol (i.e., 8) (Table 2, entry 7). However, yield. First, we demethylated the ether moiety of (+)-8

this material was almost racemic: a 5% nett inversion of the cleanly at 180 °C, by heating it with MeMgI·Et₂O

configuration had occurred.

(20 equiv.).^[69]Phenol 38 was formed in almost quantitative

Our interpretation of the absence of type-32 alcohols yield. Double esterification provided bis(p-bromobenzoate)

from the relevant product portfolios in Table 1 (i.e., en- 40. Treatment of 40 with a suspension of NaHCO₃in meth-

tries 10–20) had been that the reactions of organocopper anol selectively removed the aryl ester by transesterification

reagents with styrene oxides 29 do not tend to proceed via to give the desired monoester (i.e., 39).

benzyl cations. This prospect encouraged us to try such rea-

gents for the ring-opening of epoxide 9 (Table 2, entries 8–

13). The ethyl Gilman cuprate, however, shared the weak-

ness of its butylated counterpart (Table 1, entry 11), attack- Building Block

Conversion of the Indane into the C¹–C¹¹

ing at C^αrather than at C^β(Table 2, entry 9). The ethyl-

Having achieved the synthesis of indane-substituted

containing “higher order cyanocuprate”^[61]showed a 77:23

alcohol (+)-(S)-8 with 95% ee, our synthetic design

bias in favor of the required β-attack (Table 2, entry 10).

(Scheme 1) called for its subjection to a Birch reduction/

The ethyl-containing “lower order cyanocuprate” remained

ozonolysis sequence in order to release the hydroxy-triketo-

inert when it was exposed to epoxide 9 in THF solution at

carboxylic ester 4. Scheme 6 shows how this transformation

–20 °C (Table 2, entry 11). However, the same cuprate re-

was executed. Deferring a detailed discussion to the next

acted smoothly with epoxide 9 (99.6% ee) in diethyl ether

paragraph, it is first worthwhile to appreciate the final

solution at 0 °C (Table 2, entry 13). The regioselectivity for

Scheme 6. Realization of the Birch reduction/ozonolysis strategy

Scheme 5. Conversion of (+)-8 into p-bromobenzoate (S)-39. Con-

“8Ǟ4” shown in Scheme 1. Reagents and conditions: a) Li

figuration-proving ORTEP plot of the unit cell of an X-ray struc-

(12 equiv.), liq. NH₃/THF/tBuOH (6:1:1), –33 °C, 1 h; addition of

ture analysis of a single crystal of (S)-39 at 173 K.^[67]Reagents and NH₄^–OAc; extractive work-up; 95%. b) Stream of O₃in O₂,

+

conditions: a) MeMgI (3 m in Et₂O, 20 equiv.), neat, 180 °C, 30 min;

94%. b) p-Bromobenzoyl chloride (3.0 equiv.), pyridine (as sol-

MeOH/CH₂Cl₂(2:1), –78 °C, 45 min; PPh₃(2.2 equiv.), room

temp., 15 h; the product was not purified. c) Camphorsulfonic acid

vent), room temp., 24 h; 64%. c) NaHCO₃(3.0 equiv.), MeOH/ (20 mol-%), THF, room temp., 22 h; addition of Na₂SO₄(anhy-

THF (1:1), reflux, 3 h; 62%.

drous), 4 h; 50% over the two steps from 5.

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6909

L. Nachbauer, R. Brückner

FULL PAPER

product shown in Scheme 6. It is dihydropyranone-contain-

ing β-keto ester anhydro-4. We obtained it almost effort-

lessly from 4, which had been the key compound in our

synthetic plan (Scheme 1). The advantage of dealing with

anhydro-4 from then on rather than dealing with 4 was that

anhydro-4 protected the former C⁵=O group as an enol

ether. Moreover, anhydro-4 attenuated the electrophilicity of

the former C³=O group because it was integrated into a

push-pull olefin. The unaltered C⁹=O group of anhydro-4

was part of a β-keto ester moiety. Thus only C⁹=O, and no

other structural motif, was susceptible to a Noyori hydro-

genation,^[70]as Scheme 7 shows. Thereafter, the C³=O moi-

ety of the push-pull olefin substructure was reduced by a

metal hydride (Scheme 8). Scheme 9 details the concluding

steps towards synthetic equivalent 47 of C¹–C¹¹synthon 2

(Scheme 1) of rimocidin (1a).

Scheme 8. Stereoselective carbonyl group reduction II: Luche re-

duction of spiroketal ketones 43 and epi-43.^[84]Reagents and condi-

tions: a) NaBH₄(1.1 equiv.), CeCl₃·7H₂O (1.1 equiv.), MeOH,

–78 °C to room temp., 2 h; yields: 84% (trans-44 + cis-epi-44) sepa-

rated from 10% (cis-44).

Scheme 7. Stereoselective carbonyl group reduction I: Noyori hy-

drogenation of β-keto ester anhydro-4 and spiroketal ketone forma-

+

tion.^[75]Reagents and conditions: a) H₂(4.0 bar), Me₂NH₂

–

[RuCl(S)-(BINAP)]₂(μ-Cl)₃(0.5 mol-%), EtOH, room temp., 16 h;

42:43 = 80:20 in the crude product; 67% 42 after flash chromatog-

raphy on silica gel (43 was not isolated).^[35]b) Same as (a), then H₂

replaced by N₂(1 bar), room temp., 72 h; 68:32 mixture of 42 and

43. c) Camphorsulfonic acid (5 mol-%), CHCl₃, room temp., 16 h;

90:10 mixture of 43 and epi-43, 66% over the two steps.

Ozonolysis of Birch reduction product 5 seemed frustrat-

ing (Scheme 6). According to TLC and H NMR analyses,

a plethora of products formed in the reaction, and they did

not include the desired hydroxy-triketocarboxylic ester (i.e.

1

The Birch reduction of indane (+)-(S)-8 was performed 4). Mindful of the tendency^[26,28,29]of polycarbonyl com-

under conditions that we deemed typical for anisols,^[23]i.e., pounds to form ketals in an uncontrolled fashion, we tried

using lithium^[71]in a mixture of liquid ammonia, THF, and hard to effect the C=C cleavages in 5 step by step. We moni-

tert-butanol^[72](Scheme 6). Without purification, cy- tored the ozonolysis by TLC, performed it in the presence

clohexa-1,4-diene 5 was isolated in 95% yield. It had to of the Sudan dyes Solvent Black 3, Solvent Red 19, and

be handled under an atmosphere of argon because of its Solvent Red 23 (in decreasing order of reactivity towards

pronounced propensity to rearomatize (Ǟ8). Bearing this O₃),^[73]or added 2-methylbut-2-ene or 2,3-dimethylbut-2-

in mind, 5 was obtained remarkably pure, as shown by its ene to the substrate in order to stop the reaction after the

1

300 MHz H NMR spectrum in CDCl₃. This revealed four C³=C¹¹bond had been broken and while the C⁵=C⁹bond

allylic protons (δ ≈ 2.3 ppm) and four bisallylic protons (δ was still intact. This was the order of reactivity that would

= 2.57 and 2.84 ppm) as required, and there was no indica- be favored on the basis that “the methoxy-substituted

tion of any sizeable amounts of contaminants.

C³=C¹¹bond reacts first because it is more electron rich”.

6910

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

The opposite expectation would have been that “the tetra-

Spiroketal formation could be driven to completion only

substituted C⁵=C⁹bond reacts first because it is less steri- when we isolated the mentioned 68:32 mixture of 42 and 43,

[81]

cally hindered”. In fact, there seemed to be no overall effect, and restarted the cyclization 42Ǟ43 once more. ZnBr₂

as we never isolated a cyclopentene.

(1.2 equiv., CH₂Cl₂, room temp., 20 h) in CH₂Cl₂,

Eventually, we returned to the initially attempted “com- BF₃·Et₂O (1.2 equiv., room temp., CH₂Cl₂, 16 h) in

plete ozonolysis” of Birch reduction product (+)-(S)-8 CH₂Cl₂, or pyridinium p-toluenesulfonate in CDCl₃were

(Scheme 6). It was carried out in MeOH/CH₂Cl₂(2:1) at ineffective for this cyclization. A 5% solution of concd. HCl

–78 °C. The surmised diastereomeric mixture of secondary in THF/CH₃CN (1:1)^[82]effected it at room temp. within 2 h

ozonides 41 was destroyed by the addition of PPh₃, and not (84% yield). Unfortunately, this occurred with concomitant

Me₂S, which did not bring about a complete conversion. epimer formation – 43:epi-43 = 86:14 – and the epimers

1

Reevaluating the H NMR spectrum of the resulting mix- were inseparable by flash chromatography on silica gel.^[35]

ture, we ignored the apparent absence of hydroxy-triketo- The best 43:epi-43 ratio (90:10) was obtained when a chlo-

carboxylic ester 4, and wondered whether it showed instead roform solution of the 42:43 mixture was treated with cam-

the various plausible isomers of 4 – e.g. enol-4, cyclo-4, and/ phorsulfonic acid. The yield of 43/epi-43 then was 66% over

or spiro-4 and their diastereomers. A mass spectrum (EI, the two steps.

70 eV) of the mixture showed a 33% peak at m/z = 268.1,

The three-dimensional structures of spiroketal epimers

which is the value calculated for C₁₄H₂₀O₅⁺. This was en- 43 and epi-43 were resolved mainly through analysis of the

couraging, since the latter might be the [M – H₂O]⁺peak vicinal H,H coupling constants within the tetrahydropyran-

formed from the molecular ion M⁺of any of the isomers 4, 4-one (as opposed to tetrahydropyran) moieties (500 MHz,

enol-4, cyclo-4, or spiro-4. This interpretation was corrobo- CDCl₃; Scheme 7, bottom). Such an analysis was feasible,

rated when we managed to converge the mixture into a sin- even though none of the spiroketals was isolated pure. Our

gle product, namely dihydropyranone-containing β-keto es- configurational proofs consisted of pinpointing axially ori-

ter anhydro-4. This compound was formed in 50% overall entated protons – those protons showing diaxial couplings

yield from 5 when the ozonolysate was reduced with PPh₃(Ͼ 10 Hz). This let us assign the major spiroketal (i.e., 43)

and then exposed in THF solution to camphorsulfonic acid as an equatorially ethylated tetrahydropyran-4-one and the

for 1 d. Anhydro-4 is a dihydropyranone-containing β-keto minor spiroketal (i.e., epi-43) as an axially ethylated tetra-

ester. It was identified by singlets due to the O=C–CH₂– hydropyran-4-one. The configuration of the spirocenter in

C=O unit (δ = 3.45 ppm) and the olefinic proton of the the major spiroketal 43 followed from the equatorial orien-

1

pyranone moiety (δ = 5.27 ppm) in its H NMR spectrum tation of its CH₂CO₂Me group [which, in turn, was de-

(500 MHz, CDCl₃).^[74]

duced from the magnitudes of the “endocyclic” vicinal cou-

The β-keto ester moiety of dihydropyranone anhydro-4 pling constants (J_axial,axial= 11.5 Hz; J_{axial,equatorial}

was subjected to a Noyori hydrogenation^[70]in ethanol. 2.2 Hz) of the neighboring ring proton (9-H; cf. Scheme 7)].

=

+

–

Using Me₂NH₂[RuCl(S)-(BINAP)]₂(μ-Cl)₃as a cata- An NOE observed in the major spiroketal 43 confirmed

lyst,^[76]4 bar hydrogen pressure was sufficient to effect com- this conclusion (cf. Scheme 7). Because of signal overlap,

plete conversion at room temp. within 16 h (Scheme 7). the configuration of the spirocenter in the minor spiroketal

Flash chromatography on silica gel^[35]gave S-configured^[77]epi-43 was not proved.^[83]

β-hydroxy ester 42 in 67% yield. This was a lower yield

The next step was the reduction of the inseparable 90:10

than expected.^[78]We blamed this discrepancy on having mixture of spiroketal tetrahydropyranones 43 and epi-43

lost some 42 because it had cyclized under the reaction con- (Scheme 8). The desired configuration of the stereocenter

ditions^[79]to give spiroketal 43. This compound had been formed at C³would result from reduction of the major iso-

detected as a pure epimer (i.e., free from epi-43) as a minor mer (i.e., 43) by axial attack. Such a selectivity is known

1

constituent in the H NMR spectrum (500 MHz, CDCl₃) using NaBH₄,^[85]SmI₂,^[86]or potassium dissolving in NH₃/

of the crude hydrogenation product (42:43 = 80:20).

MeOH^[87]as a reductant. We used NaBH₄/CeCl₃^[88]and ob-

Spiroketal 43 appeared more attractive as a synthetic in- tained three out of the maximum of four possible dia-

termediate than its precursor, pyranone-containing β-hy- stereomers. The major spiroketal (i.e., 43) was reduced to

droxy ester 42. This is because the forthcoming reduction give spiroketal alcohols trans-44 and cis-44. The minor spi-

of the C³=O group would give a more stable alcohol start- roketal (i.e., epi-43) gave the spiroketal alcohol cis-epi-44,

ing from 43 (Ǟ44; cf. Scheme 8) than starting from 42.^[80]selectively. Flash chromatography allowed the separation of

For this reason, we tried to drive the β-hydroxy ester 42 these products into an 89:11 mixture of trans-44 and cis-

deliberately (cf. ref.^[79]) towards spiroketalization (Ǟ43). epi-44 (84% yield), and pure cis-44 (10% yield).

For instance, we left the crude hydrogenation product(s) 42

Diastereomer cis-44 was analyzed by ¹H NMR spec-

(+ 43) for three extra days [Scheme 7, experiment (b) vs. (a)] troscopy (Scheme 8) and X-ray crystallography (Fig-

1

in the HCl-containing^[79]solvent. The 300 MHz H NMR ure 1).^[67]Its 3D structure originated (1) from the configura-

spectrum of the crude product mixture revealed a 68:32 ra- tion of spiroketal tetrahydropyranone 43, in which the ethyl

tio of β-hydroxy ester 42 and spiroketal 43 (without any group at C²was equatorial, and (2) from an equatorial at-

epi-43). Exposing the crude hydrogenation product(s) 42 tack of hydride.

(+ 43) to the ethanolic HCl^[79]for 5 d allowed the yield of

formation of 43 to reach 68%.

Spiroketal alcohol trans-44 must have originated from

spiroketal tetrahydropyranone 43 for mass balance reasons.

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6911

L. Nachbauer, R. Brückner

FULL PAPER

ization conditions [ethane-1,2-dithiol (10 equiv.), BF₃·Et₂O

(10 equiv.), –40 °C, 7 h] than before allowed the conversion

of cis-epi-44 into epi-45 without affecting trans-44 at all.

Compound trans-44 was reisolated free of its epimer in

quantitative yield. Subjecting the reisolated trans-44 to the

earlier described “harsher” transthioketalization conditions

[ethane-1,2-dithiol (100 equiv.), BF₃·Et₂O (60 equiv.),

–40 °C, 3 d] gave a single thioketal, polyketide 45 (81%

yield). Its alcohol functionalities were silylated with

TBSOTf and 2,6-lutidine (Ǟ46, 85% yield).^[94,95]Re-

duction of the ester group with DiBAH (1.2 equiv.) in

CH₂Cl₂yielded aldehyde 48 without overreduction so

cleanly (90% yield) that purification by flash chromatog-

raphy^[35]could be omitted.

Figure 1. ORTEP plot of an X-ray structure analysis of a single

crystal of spiroketal cis-44 (recorded at 173 K).^[67]The axial OH

group is stabilized by two hydrogen bonds.

This is because the former was the major product and the

latter the predominant component of the starting material.

Hence, like tetrahydropyranone precursor 43, spiroketal

alcohol trans-44 contains an equatorial ethyl group at C².

The OH group at C³of trans-44 is also equatorial, because

3-H (δ_{in C6D6}= 0.96 ppm) is axial, as deduced from the

magnitude of the coupling constant J_3,4of 11.1 Hz,^[89]

which indicates a diaxial coupling (Scheme 8). The axial

orientation of 3-H in the major reduction product (i.e.,

trans-44) of tetrahydropyranone 43 gives evidence for the

preponderance of an equatorial rather than an axial hydride

attack.

Having identified cis-44 and trans-44 as reduction prod-

ucts with equatorial ethyl groups, and thus products of the

reduction of spiroketal tetrahydropyranone 43, it was clear

that the third spiroketal alcohol shown in Scheme 8

stemmed from the minor tetrahydropyranone precursor

(i.e., epi-43) and therefore contained an axial ethyl group

(and an equatorial 2-H bond^[90]). The OH group in this

spiroketal alcohol is equatorial, and 3-H is axial. This was

concluded from the fact that the high-field proton at C-

4 (δ_{in C6D6}= 1.41 ppm) shows a diaxial coupling constant

Scheme 9. Separation of reduction products trans-44 and cis-epi-

44 (89:11 mixture) by an epimer-selective trans(thio)ketalization.

Conversion of trans-44 into C¹–C¹¹building block 47. Reagents

and conditions: a) HS(CH₂)₂SH (100 equiv.), BF₃·Et₂O (60 equiv.),

–40 °C, 72 h; 63% of an 89:11 45:epi-45 mixture. b) HS(CH₂)₂SH

(10 equiv.), BF₃·Et₂O (10 equiv.), –40 °C, 7 h; 98% trans-44 reiso-

lated; the yield of epi-45 was not determined. c) Same conditions as

(a); 81%. d) TBSOTf (3.5 equiv.), 2,6-lutidine (7.0 equiv.), CH₂Cl₂,

–78 °C to room temp., 30 min, 85%. e) DiBAH (1.2 equiv.),

CH₂Cl₂, –78 °C, 30 min; 90%. f) HS(CH₂)₃SH (3.5 equiv.), LiClO₄

(3.0 equiv.), Et₂O, room temp., 24 h; 63%. TBS = tert-butyldime-

thylsilyl; DiBAH = diisobutylaluminum hydride.

[91]

(11.7 Hz) for J_4,3

.

Thus, we assigned stereostructure cis-

epi-44 to the third spiroketal alcohol shown in Scheme 8.

In spiroketal alcohol trans-44, the stereocenters at C², C³,

and C⁹were correctly installed. The refunctionalization of

the polyketide chain by breaking the spiroketal open was

effected by an O,OǞS,S transketalization with ethane-1,2-

dithiol^[92](Scheme 9). This reaction required a large excess

of both the dithiol (100 equiv.) and BF₃·Et₂O (60 equiv.) to

reach completion within 3 d at –40 °C.^[93]An inseparable

89:11 mixture of epimers 45 and epi-45 resulted (63%

yield). The minor spiroketal (i.e., cis-epi-44) reacted more

The formation of 1,3-dithiane 47 from aldehyde 48 re-

quickly, with its conversion being complete after 24 h. The quired

a careful search for appropriate conditions

rate order of formation of epi-45 Ͼ 45 was plausible, be- (Scheme 9). This was because a Lewis acid was required as

cause cis-epi-44 is stabilized by one anomeric effect and a catalyst, but this risked cleaving the silyl ethers.^[96]Indeed,

[97]

trans-44 by two. The difference in their transthioketalization the use of BF₃·Et₂O or TiCl₄

at –78 °C resulted in the

rates opened the possibility of differentiating the sprioketal decomposition of substrate 48. Neither was thioacetali-

[98]

alcohols by effecting a partial transthioketalization, i.e., a zation feasible using SOCl₂/SiO₂

nor trimethylsilyl-pro-

“kinetic resolution” of sorts. Indeed, milder transthioketal- tected propane-1,3-dithiol/ZnI₂without jeopardizing the

6912

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

silyl ether moieties.^[99]Finally, propane-1,3-dithiol in con- epi-44 (Ǟepi-45), allowing the reisolation of spiroketal

junction with the mildly Lewis acidic LiClO₄in Et₂O^[100]alcohol trans-44 as a pure compound. It was then trans-

delivered the desired dithiane (i.e., 47) without by-products thioketalized (Ǟ45) itself, which set the stage for DiBAH

in 63% yield. Compound 47 is a protected C¹–C¹¹fragment reduction and dithiane formation. This reaction had to oc-

of aglycon 1b of rimocidin (1a).

cur in the presence of three TBS ether groups, and proved

For an experimentalist, making dithianes (here: 47) from to be possible by using 1,3-propanedithiol under LiClO₄

carbonyl compounds (here: 48) and propane-1,3-dithiol is catalysis. This step delivered the C¹–C¹¹building block (i.e.,

not a well-liked transformation because of the nauseous 47) of rimocidin (1a) and its aglycon 1b.

and intense smell of the thiol. We worked up the reaction

for the synthesis of dithiane 47 using a procedure that

should be welcomed by many others, because it scavenges

the propane-1,3-dithiol and so keeps it from evaporating.

In the case at hand, this innovation was not for olfactory

Experimental Section

concerns but simply because it was troublesome to separate

excess propane-1,3-dithiol (in the earlier fractions) from the

General Information: Reactions were performed under N₂in glass-

ware that had been freshly dried with a heat-gun under vacuum.

THF was freshly distilled from K, Et₂O was freshly distilled from

desired dithiane (i.e., 47; in the later fractions) by flash

chromatography on silica gel.^[35]We avoided this problem Na/K, CH₂Cl₂and NEt₃were distilled from CaH₂. Petroleum ether

refers to distillates with a boiling point of 30–50 °C. Products were

purified by flash chromatography^[35]on Merck silica gel 60 (0.040–

0.063 mm). Fraction numbers are given (#). Unless otherwise

stated, yields refer to analytically pure samples. ¹H NMR spec-

troscopy [TMS (tetramethylsilane; δ = 0.00 ppm) as internal stan-

dard in CDCl₃; C₆HD₅(δ = 7.16 ppm) as internal standard in

C₆D₆]: Varian Mercury VX 300, Bruker Avance 400, and Bruker

DRX 500. ¹³C NMR spectroscopy [TMS (δ = 0.00 ppm) as internal

standard in CDCl₃; CHD₅(δ = 128.06 ppm) as internal standard

in C₆D₆]: Bruker Avance 400 and Bruker DRX 500. Assignments

of ¹H NMR and ¹³C NMR resonances refer to the IUPAC no-

menclature except within substituents (where primed numbers are

used). MS: Dr. J. Wörth, C. Warth, Institut für Organische Chemie

und Biochemie, University of Freiburg. Combustion analyses: F.

Tönnies and A. Siegel, Institut für Organische Chemie und Bioche-

mie, Universität Freiburg. IR spectra: Perkin–Elmer Paragon 1000.

Optical rotations were measured with a Perkin–Elmer polarimeter

341 at 589 nm and 20 °C, and were calculated by the Drude equa-

tion {[α]_D= (α_expϫ100)/(cϫd)}; rotational values are the average

of five measurements of α_expin a given solution of the respective

sample. Melting points were measured on a Dr. Tottoli apparatus

by liberating the crude dithiane 47 from excess propane-1,3-

dithiol. We did so by pouring the ethereal reaction mixture

into a separatory funnel and shaking it vigorously with an

aqueous CuSO₄solution, which had been buffered to pH ≈

5 by NaHCO₃. The thiol and Cu^IIformed a yellow precipi-

tate, which was easily filtered off. This work-up procedure

is likely to be unknown to many in the synthetic community

if not all, and it should be broadly welcomed as it can re-

duce a classic malodor.

Conclusions

Starting from 5-indanol (21), we synthesized a fully func-

tionalized and differentially protected C¹–C¹¹building

block 47 for the polyol-polyene macrolide rimocidin (1a).

The total yield was 2.3% over 16 steps, the average yield

79% per step. 5-Methoxyindane (22) was acylated regiose-

lectively by using a nitrile as both solvent and electrophile;

this was the only way to preserving the methoxy group. The (Büchi). The ee values were determined by chiral HPLC with a

first stereocenter (at C²) was introduced with 99.6% ee by a

Chiralcel OD-H (0.46ϫ25 cm, Daicel Chemical Ind. Ltd.) column

by G. Fehrenbach, Institut für Organische Chemie und Biochemie,

Corey–Itsuno reduction of acetophenone 24. The resulting

chlorohydrin (i.e., 25) cyclized under basic conditions to

give styrene oxide 9. This compound was ring-opened

[Ǟ(+)-(S)-8] with high chemo-, regio-, and stereoselectivity

Universität Freiburg. X-ray data for 23 and cis-52^[67]were recorded

at 173 K by Dr. M. Keller, Institut für Organische Chemie und

Biochemie, Universität Freiburg.

only by the lower-order cyanocuprate EtCu(CN)Li in Et₂O

solution, but not by the many other organometallics we

tried. Since ring-opening product 8 was an indane, too, it

was susceptible to a Birch reduction. A double ozonolysis

of the resulting cyclopentane-annulated cyclohexadiene

(i.e., 5) delivered a hydroxy-triketo ester 4. It cyclocon-

densed to give β-keto ester-substituted dihydropyranone

anhydro-4. The C³=O, C⁵=O, and C⁹=O groups of precur-

sor 4 were differentiated in such a way that a Noyori hydro-

genation of the C⁹=O group followed by a Luche reduction

of the C³=O group completed the stereogenic centers in the

spiroketal alcohol trans-44. The latter compound could not

be freed from a small amount of its undesired epimer cis-

epi-44 by flash chromatography,^[35]but the epimers were

separated by an isomer-selective transthioketalization with

Methyl (R)-6-(3-Ethyl-4-oxo-3,4-dihydro-2H-pyran-6-yl)-3-oxohex-

anoate (anhydro-4) Equilibrating with Methyl (R,Z)-6-(3-Ethyl-4-

oxo-3,4-dihydro-2H-pyran-6-yl)-3-hydroxyhex-2-enoate (enol-anhy-

dro 4):

Dihydroarene 5 (2.42 g, 10.9 mmol, 1.0 equiv.) was dissolved in

MeOH/CH₂Cl₂(3:1, 80 mL). A stream of O₃/O₂was passed

through the solution at –78 °C for 45 min, until a blue color per-

sisted. Excess O₃was removed by passing a stream of N₂through

1,2-ethanedithiol/BF₃·Et₂O. This reaction affected only cis- the reaction mixture at –78 °C for 20 min. After addition of PPh₃

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6913

L. Nachbauer, R. Brückner

FULL PAPER

(6.28 g, 23.9 mmol, 2.2 equiv.), the mixture was allowed to reach (S)-2-(6-Methoxyindan-5-yl)butan-1-ol (8):

room temp. and stirred for 1 h. The solvent was evaporated under

reduced pressure, the residue was dissolved in THF (60 mL), and

camphorsulfonic acid (506 mg, 2.18 mmol, 0.2 equiv.) was added.

After stirring at room temp. for 18 h, dry Na₂SO₄(1.55 g,

10.9 mmol, 1 equiv.) was added. The reaction was continued for

4 h. NEt₃(1.5 mL, 10.9 mmol, 1.0 equiv.) was added, and all vola-

tiles were evaporated under reduced pressure. Flash chromatog-

EtLi (1.74 m in nBu₂O, commercially available at ACROS, 35 mL,

raphy^[35](8 cm, fraction volume 100 mL, Et₂O) of the residue gave

a 91:9* mixture of anhydro-4 and enol-anhydro-4 as a colorless oil

(#11–17, 1.46 g, 50%). * This ratio was determined from the ¹H

NMR integrals of the signals at δ = 5.27 (s, 1 H, 5Ј-H, anhydro-4)

and 5.28 (s, 1 H, 5Ј-H, enol-anhydro-4) ppm. [α]²_D⁰= –4.7 (c = 1.30

61.1 mmol, 1.7 equiv.) was added at –40 °C to a stirred suspension

of CuCN (5.47 g, 61.1 mmol, 1.7 equiv.) in Et₂O (180 mL). After

the addition was complete, stirring was continued at –20 °C for

30 min. The black solution was cooled to –40 °C. The neat epoxide

9 (6.84 g, 36.0 mmol, 1.0 equiv.) was added dropwise. The resulting

mixture was stirred at 0 °C for 4 h, and at room temp. for 16 h.

Buffer solution [prepared from NH₄Cl (satd. aq.), NH₃(25% aq.),

and H₂O (4:3:1); 120 mL] was added slowly. The resulting mixture

was stirred vigorously for 2 h at room temp while being exposed to

the ambient atmosphere. After extraction with Et₂O (3ϫ 50 mL),

the combined organic extracts were washed with brine (50 mL),

dried with Na₂SO₄, and filtered, and the solvents were evaporated

in vacuo. The residue was purified by flash chromatography^[35]

(8 cm, fraction volume 100 mL, cyclohexane/EtOAc, 7:1) to yield

8 as a colorless oil (#10–20, 6.04 g, 76%). The ee of (S)-8 was

determined by chiral HPLC as specified in ref.^[63][α]²_D⁰= +10.0 (c

1

in CHCl₃, 10 cm). H NMR (499.6 MHz, CDCl₃/TMS): anhydro-

4: δ = 0.98 (dd, J_2ЈЈ,1ЈЈA= J_2ЈЈ,1ЈЈB= 7.5 Hz, 3 H, 2ЈЈ-H₃), 1.44 (ddq,

²J_{1ЈЈA,1ЈЈB}= 14.4, J_1ЈЈA,5Ј= 7.6, J_1ЈЈA,2ЈЈ= 7.2 Hz, 1 H, 1ЈЈ-H^A), 1.82

(ddq, ²J_{1ЈЈB,1ЈЈA}= 13.8, J_1ЈЈB,2ЈЈ= 7.6, J_1ЈЈB,5Ј= 5.6 Hz, 1ЈЈ-H^B), 1.88

(tt, J_5,4= J_5,6= 7.3 Hz, 2 H, 5-H₂), 2.26 (t, J_6,5= 7.3 Hz, 2 H, 6-

H₂), overlapping with 2.23–2.30 (m, 1 H, 3Ј-H), 2.60 (t, J_4,5

=

7.1 Hz, 2 H, 4-H₂), 3.45 (s, 2 H, 2-H₂), 3.74 (s, 3 H, OCH₃), 4.21

(dd, ²J_6ЈA,6ЈB= 11.4, J_6ЈA,5Ј= 8.4 Hz, 1 H, 6Ј-H^A), 4.44 (dd, ²J_6ЈB,6ЈA

= 11.4, J_6ЈB,5Ј= 4.75 Hz, 1 H, 6Ј-H^B), 5.27 (s, 1 H, 5Ј-H) ppm;

enol-anhydro-4: δ = 5.00 (s, 1 H, 2-H), 5.27 (s, 1 H, 5Ј-H), 12.03 (s,

1 H, 3-OH) ppm. ¹³C NMR (125.6 MHz, CDCl₃/TMS): δ = 11.5

(C-2ЈЈ), 19.97 (C-5), 20.11 (C-1ЈЈ), 33.54 (C-6), 41.69 (C-4), 45.39

(C-3Ј), 49.06 (C-2), 52.44 (OCH₃), 71.25 (C-2Ј), 104.05 (C-5Ј),

167.48 (C-1), 175.61 (C-6Ј), 195.10 (C-4Ј), 201.63 (C-3) ppm. IR

1

= 1.04 in CHCl₃, 10 cm). H NMR (400.1 MHz, CDCl₃/TMS): δ

= 0.86 (dd, J_4,3A= J_4,3B= 7.4 Hz, 3 H, 4-H₃), AB signal (δ_A

1.61, δ_B= 1.74, J_AB= 13.5 Hz, in addition split by J_A,4= J_B,4

7.4, J_A,2= 8.7, J_B,2= 6.1 Hz, 2 H, 3-H₂), 2.07 (tt, J_2Ј,1Ј= J_2Ј,3Ј

=

(film): ν = 3465, 2965, 2880, 1745, 1715, 1665, 1610, 1450, 1440,

˜

1405, 1365, 1325, 1205, 1170, 1110, 1010, 900, 825 cm^–1. C₁₄H₂₀O₅

(268.31): calcd. C 62.67, H 7.51; found C 62.41, H 7.65.

7.4 Hz, 2 H, 2Ј-H₂), 2.84 (t, J_3Ј,2Ј= 7.7 Hz, 2 H, 3Ј-H₂),* 2.88 (t,

J_1Ј,2Ј= 7.7 Hz, 2 H, 1Ј-H₂),* 3.18 (dddd, J_2,3A= 8.7, J_2,1A= 7.0,

J_2,1B= J_2,3B= 6.0 Hz, 1 H, 2-H), AB signal (δ_A= 3.72, δ_B= 3.75,

J_AB= 10.5 Hz, in addition split by J_A,2= 7.1, J_B,2= 5.8 Hz, 2 H,

1-H₂), 3.79 (s, 3 H, OCH₃), 6.79 (s, 1 H, 7Ј-H),* 7.01 (s, 1 H, 4Ј-

H)* ppm. ¹³C NMR (100.6 MHz, CDCl₃/TMS): δ = 12.17 (C-4),

24.07 (C-3), 25.67 (C-2Ј), 32.26 (C-3)Ј,* 33.21 (C-1Ј),* 42.92 (C-2),

55.76 (OCH₃), 66.67 (C-1), 107.34 (C-7Ј),* 132.53 (C-4Ј),* 128.24

(C-6Ј), 136.04 (C-9Ј), 143.26 (C-8Ј), 156.80 (C-5Ј) ppm; * these as-

signments were corroborated by an HMBC experiment. IR (film):

(S)-2-(6-Methoxy-4,7-dihydroindan-5-yl)butan-1-ol (5):

ν = 3365, 2955, 2870, 1615, 1490, 1465, 1415, 1300, 1260, 1195,

˜

Ammonia (approx. 300 mL) was condensed into a three-necked

round-bottomed flask at –78 °C. Small pieces (approx.

2ϫ2ϫ2 mm) of lithium (2.00 g, 286 mmol, 12 equiv.) were added

under vigorous stirring. After they had dissolved, a solution of 8

(5.25 g, 23.8 mmol, 1.0 equiv.) in THF/tBuOH (1:1, 50 mL) was

added dropwise. The resulting mixture was stirred at –33 °C for

1 h. Excess reductant was destroyed by adding wet ammonium

acetate (25 g) in five portions such that the mixture became color-

less and stayed so. The ammonia was allowed to evaporate at room

temp. H₂O (100 mL) was added. The resulting mixture was ex-

tracted with Et₂O (3ϫ 50 mL). The extracts were combined and

washed with H₂O (50 mL) and brine (50 mL), dried with MgSO₄,

and evaporated under reduced pressure. A colorless oil (5.03 g,

95%) was obtained. It was stored without purification – because

the ¹H NMR spectrum showed that the sample was pure 5 – at

4 °C under an argon atmosphere. In this manner, re-aromatization

1170, 1080, 1040, 915, 840, 745 cm^–1. C₁₄H₂₀O₂(220.31): calcd. C

76.33, H 9.15; found C 76.13, H 9.36.

(R)-2-(6-Methoxyindan-5-yl)oxirane (9):

Aqueous NaOH (2 m, 130 mL, 260 mmol, 7.0 equiv.) was added at

10 °C to

a solution of chlorohydrin 25 (8.55 g, 37.7 mmol,

1.0 equiv.) in iPrOH (130 mL) while stirring. Stirring was continued

vigorously at room temp. for 1.5 h. The resulting mixture was ex-

tracted with Et₂O (2ϫ 100 mL). The combined organic extracts

were washed with H₂O (100 mL), aq. phosphate buffer (pH = 7.1,

0.5 m, 100 mL), and brine (100 mL). They were dried with Na₂SO₄,

filtered, and evaporated in vacuo. The resulting yellowish oil

(6.85 g, 96%) was not purified, since it decomposed on silica gel.

A ¹H NMR spectrum of 9 revealed no impurities. Compound 9 was

1

was avoided until the compound was submitted to ozonolysis. H

NMR (400.1 MHz, CDCl₃/TMS): δ = 0.88 (dd, J_4,3A= J_4,3B

=

7.4 Hz, 3 H, 4-H₃), AB signal (δ_A= 1.31, δ_B= 1.41, J_AB= 13.4 Hz,

in addition split by J_A,2= 9.5, J_A,4= 7.3, J_B,2= 5.8, J_B,4= 7.6 Hz,

2 H, 3-H₂), 1.60 (br. s, 1 H, 1-OH), 1.90 (tt, J_2Ј,1Ј= 8.1,* J_2Ј,3Ј

=

1

6.8 Hz,* 2 H, 2Ј-H₂), 2.23–2.34 (m, 4 H, 1Ј-H₂, 3Ј-H₂), 2.57 and

2.84 (2 m, 4 H, 4Ј-H₂, 7Ј-H₂), 3.00 (dddd, J_2,3A= J_2,1A= 9.5, J_2,3B

= J_2,1B= 5.5 Hz, 1 H, 2-H), 3.44–3.62 (m, 2 H, 1-H₂), overlapping

with 3.54 (s, 3 H, OCH₃) ppm; * values interchangeable.

stored at 4 °C, where it crystallized slowly. H NMR (300.1 MHz,

CDCl₃/TMS): δ = 2.06 (tt, J_2Ј,1Ј= J_2Ј,3Ј= 7.4 Hz, 2 H, 2Ј-H₂), 2.70

(dd, ²J_3cis,3trans= 5.7, J_3cis,2= 2.7 Hz, 1 H, 3-H^cis), 2.81 and 2.89 (2

t, J_1Ј,2Ј= J_3Ј,2Ј= 7.4 Hz, 4 H, 1-H₂, 3Ј-H₂), 3.84 (s, 3 H, OCH₃),

6914

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

3.11 (dd, ²J_3trans,3cis= 5.7, J_3trans,2= 4.1 Hz, 1 H, 3-H^trans), 4.17 (dd, with H₂O (2ϫ 100 mL) and brine (2ϫ 100 mL), dried with

J_2,3trans= 4.0, J_2,3cis= 2.8 Hz, 1 H, 2-H^cis) ppm.

Na₂SO₄, and filtered, and the solvents were evaporated in vacuo.

The residue was recrystallized from n-hexane to yield 24 as a

brownish crystalline solid (9.33 g, 60%), m.p. 62–64 °C. H NMR

5-Methoxyindane (22):^[101]

1

(400.1 MHz, CDCl₃/TMS): δ = 2.10 (tt, J_2Ј,1Ј= J_2Ј,3Ј= 7.5 Hz, 2

H, 2Ј-H₂), 2.85 and 2.93 (2 t, J_1Ј,2Ј= J_3,2= 7.4 Hz, 4 H, 1Ј-H₂, 3Ј-

H₂), 3.91 (s, 3 H, OCH₃), 4.77 (s, 2 H, 1Ј-H₂), 6.86 (s, 1 H, 7Ј-H),

7.71 (s, 1 H, 4Ј-H) ppm. ¹³C NMR (100.6 MHz, CDCl₃/TMS): δ

= 25.69 (C-2Ј), 31.74 and 33.70 (C-1Ј, C-3Ј), 51.27 (C-2), 55.82

(OCH₃), 107.76 (C-7Ј), 123.24 (C-5Ј), 126.59 (C-4Ј), 136.93 (C-9Ј),

Aqueous KOH [KOH (100 g) in H₂O (100 mL), 1.55 mol,

4.0 equiv.] was added to a solution of 5-indanol (21; 50.7 g,

378 mmol, 1.0 equiv.) in acetone (600 mL). Me₂SO₄(41 mL, 54.7 g,

453 mmol, 1.2 equiv.) was added dropwise over 1 h while stirring,

which resulted in the acetone refluxing gently. A precipitate of

K₂SO₄was filtered off. The aqueous phase was separated and ex-

tracted with acetone (50 mL). The combined organic extracts were

concentrated in vacuo. Fractional distillation of the residue (65–

67 °C, 1 mbar) yielded the title compound as a yellowish oil (50.9 g,

91%). ¹H NMR (300.1 MHz, CDCl₃/TMS): δ = 2.07 (tt, J_2,1= J_2,3

= 7.4 Hz, 2 H, 2-H₂), 2.84 and 2.88 (2 t, J_1,2= J_3,2= 7.3 Hz, 4 H,

1-H₂, 3-H₂), 3.78 (s, 3 H, OCH₃), 6.69 (dd, J_6,7= 8.3, ⁴J_6,4= 2.5 Hz,

1 H, 6-H), 6.78 (d, ⁴J_4,6= 2.4 Hz, 1 H, 4-H), 7.10 (d, J_7,6= 8.1 Hz,

1 H, 7-H) ppm.

152.67 (C-8Ј), 158.59 (C-6Ј), 192.24 (C-1) ppm. IR (film): ν = 2950,

˜

2840, 1680, 1615, 1570, 1485, 1465, 1415, 1255, 1200, 1165, 1145,

1100, 1025, 875, 845, 790, 745 cm^–1. C₁₂H₁₃ClO₂(224.68): calcd. C

64.15, H 5.83; found C 64.10, H 5.99.

(R)-2-Chloro-1-(6-methoxyindan-5-yl)ethanol (25):

BH₃·THF complex (1.0 m in THF, 50 mL, 49.8 mmol, 1.0 equiv.)

was added to a solution of (R)-5-(hydroxydiphenylmethyl)pyrroli-

dine-2-one^[38](266 mg, 1.00 mmol, 2 mol-%) in THF (80 mL) while

stirring at room temp. After stirring for a further 5 min, a solution

of ketone 24 (11.2 g, 49.8 mmol, 1.0 equiv.) in THF (80 mL) was

added dropwise over 3 h. After allowing the reduction to proceed

for a further 2 h, excess reductant was destroyed at 0 °C by the

slow addition of aq. HCl (2 m, 100 mL). The resulting mixture was

extracted with Et₂O (2ϫ 100 mL). The combined organic extracts

were washed with H₂O (100 mL), satd. aq. NaHCO₃(100 mL), aq.

phosphate buffer (pH = 7.1, 0.5 m, 100 mL), and brine (100 mL).

They were then dried with Na₂SO₄and filtered, and the solvents

were evaporated in vacuo. The resulting residue was recrystallized

from n-hexane. This yielded 25 as a white crystalline solid (8.55 g,

76%), m.p. 92–94 °C. The ee of 25 was determined by chiral HPLC:

Chiralcel OD-H column, n-heptane/EtOH (97:3), 0.8 mLmin^–1,

5-Iodo-6-methoxyindane (23):^[102]

A solution of iodine monochloride (3.1 mL, 60.9 mmol, 1.3 equiv.)

in AcOH (50 mL) was added dropwise to a solution of indane 22

(6.94 g, 46.8 mmol, 1.0 equiv.) in AcOH (100 mL) at room temp.

The reaction mixture was stirred at room temp. for 2 h, and then

it was extracted with petroleum ether (3ϫ 100 mL). The combined

organic extracts were carefully washed with NaHCO₃(satd. aq.;

2ϫ 100 mL), dried with Na₂SO₄, filtered, and concentrated in

vacuo. The residue was a solid, which was recrystallized from wet

EtOH to give 23 as a white crystalline solid (6.31 g, 49%, 95:5

mixture with 4-iodo-5-methoxyindane), m.p. 38–40 °C. ¹H NMR

(300.1 MHz, CDCl₃/TMS): δ = 2.07 (tt, J_2,1= 7.2,* J_2,3= 7.5 Hz,*

2 H, 2-H₂), 2.83 and 2.86 (2 t, J_1,2= J_3,2= 7.4 Hz, 4 H, 1-H₂, 3-

H₂), 3.84 (s, 3 H, OCH₃), 6.74 (s, 1 H, 7-H), 7.59 (s, 1 H, 4-H)

ppm; * assignments interchangeable.

20 °C isothermal; λ_detector= 285 nm; t_{r (R)-25}= 12.6 min, t_{r (S)-25}

=

11.6 min [(R)-25:(S)-25 = 99.8:0.2 (99.6% ee)]. [α]²_D⁰= –34.9 (c =

1

1.04 in CHCl₃, 10 cm). H NMR (400.1 MHz, CDCl₃/TMS): δ =

2.07 (tt, J_2Ј,1Ј= J_2Ј,3Ј= 7.4 Hz, 2 H, 2Ј-H₂), 2.85 and 2.89 (2 t, J_1Ј,2Ј

= J_3Ј,2Ј= 7.7 Hz, 4 H, 1Ј-H₂, 3Ј-H₂), 2.91 (d, J_1-OH,1= 2.9 Hz, 1

H, 1-OH), AB signal (δ_A= 3.62, δ_B= 3.82, J_AB= 10.9 Hz, in

addition split by J_A,1= 8.5, J_B,1= 3.6 Hz, 2 H, 2-H₂), 3.82 (s, 3 H,

OCH₃), 5.07 (ddd, J_1,2A= 8.2, J_1,2B= 4.0, J_1,1-OH= 3.7 Hz, 1 H,

1-H), 6.78 (s, 1 H, 7Ј-H), 7.26 (s, 1 H, 4Ј-H) ppm. ¹³C NMR

(100.6 MHz, CDCl₃/TMS): δ = 25.68, (C-2Ј), 31.96 and 32.13 (C-

1Ј, C-3Ј), 49.75 (C-2), 55.47 (OCH₃), 70.75 (C-1), 106.93 (C-7Ј),

122.84 (C-4Ј), 125.71 (C-5Ј), 136.14 (C-8Ј), 145.36 (C-9Ј), 155.22

2-Chloro-1-(6-methoxyindan-5-yl)ethanone (24):

(C-6Ј) ppm. IR (film): ν = 3445, 2955, 2905, 2845, 1615, 1585, 1490,

˜

1465, 1430, 1320, 1300, 1275, 1255, 1165, 1080, 1060, 1025, 915,

870, 840, 735, 700 cm^–1. C₁₂H₁₅ClO₂(226.70): calcd. C 63.57, H

6.83; found C 63.58, H 6.67.

A suspension of ZnCl₂(23 g, 858 mmol, 2.5 equiv.) in SOCl₂

(25 mL) was heated at reflux for 0.5 h. Most of the SOCl₂was

distilled off at atmospheric pressure, and the rest was removed in

vacuo (6 h). The residue of dry ZnCl₂was dissolved in chloroaceto-

nitrile (26 mL, 139 mmol, 6.0 equiv.). 5-Methoxyindane 22 (10.3 g,

69.3 mmol, 1.0 equiv.) was added. The resulting mixture was stirred

vigorously at 85 °C for 2 h, while a stream of dry HCl gas^[103]was

bubbled through. A dark-red viscous mass formed. It was cooled to

room temp. HCl (concd. aq.; 50 mL) was added dropwise (caution:

gaseous HCl was liberated). The resulting suspension was heated

at reflux for 1 h in order to hydrolyze the imine intermediate. After

cooling to room temp., the reaction mixture was extracted with

EtOAc (3ϫ 100 mL). The combined organic extracts were washed

(S)-2-(6-Hydroxyindan-5-yl)butyl 4-Bromobenzoate (39)

Cleavage of the Methyl Ether Moiety of Compound 8: Under an

atmosphere of N₂, a mixture of indane 8 (300 mg, 1.36 mmol,

1.0 equiv.) and MeMgI (3 m in Et₂O, 9.0 mL, 27 mmol, 20 equiv.)

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6915

L. Nachbauer, R. Brückner

FULL PAPER

was heated at 180 °C for 30 min. The residue was cooled to –20 °C.

H₂O (10 mL) was added slowly and with vigorous stirring. After

extraction with Et₂O (3ϫ 5 mL), the combined organic extracts

were washed with NH₄Cl (satd. aq.; 3 mL) and brine (3 mL), dried

with Na₂SO₄, and filtered, and the solvents were evaporated in

vacuo. The resulting colorless oil (262 mg, 94%) was not purified,

anhydro-4 (1.21 g, 4.51 mmol, 1.0 equiv.) in degassed EtOH

(35 mL) at room temp. The mixture was stirred for 17 h in an auto-

clave under an atmosphere of H₂(4.0 bar). The autoclave was

vented, and the reaction mixture was concentrated in vacuo. The

residue was purified by flash chromatography^[35](4 cm, fraction

volume 20 mL, cyclohexane/EtOAc, 5:8) to yield the title com-

pound as a colorless oil (#11–27, 812 mg, 67%). [α]²_D⁰= +3.8 (c =

1

since a H NMR spectrum revealed that it was essentially pure.

1

1.18 in CHCl₃, 10 cm). H NMR (400.1 MHz, CDCl₃/TMS): δ =

Formation of a Bis(4-bromobenzoate): The aforementioned residue

(50 mg, 0.24 mmol, 1.0 equiv.) was dissolved in pyridine (1 mL). p-

Bromobenzoyl chloride (160 mg, 0.727 mmol, 3.0 equiv.) was

added at room temp. After having been stirred for 24 h, the reaction

mixture was poured into aq. CuSO₄(1 m, 3 mL). The resulting mix-

ture was extracted with Et₂O (3ϫ 3 mL). The combined organic

extracts were washed with aq. CuSO₄(1 m, 3 mL) and brine (3 mL),

dried with Na₂SO₄, and filtered, and the solvents were evaporated

in vacuo. The residue, a colorless oil (88 mg, 64%), was used in the

next step without purification.

0.98 (dd, J_2ЈЈ,1ЈЈA= J_2ЈЈ,1ЈЈB= 7.5 Hz, 3 H, 2ЈЈ-H₃), 1.45 (ddq,

²J_{1ЈЈA,1ЈЈB}= 15.2, J_1ЈЈA,3Ј= J_1ЈЈA,2ЈЈ= 7.3 Hz, 1 H, 1ЈЈ-H^A), overlap-

ping with 1.42–1.59 (m, 2 H, 5-H₂), overlapping with 1.58 –1.70

(m, 1 H, 4-H^A), 1.71–1.81 (m, 1 H, 4-H^B), overlapping with 1.82

2

(ddq, J_{1ЈЈB,1ЈЈA}= 14.0, J_1ЈЈB,3Ј= 5.6, J_1ЈЈB,2ЈЈ= 7.6 Hz, 1 H, 1ЈЈ-

H^B), 2.26 (dddd, J_3Ј,1ЈЈA= 8.7, J_3Ј,2ЈA= 8.2, J_3Ј,2ЈB≈ J_3Ј,1ЈЈB≈ 4.7 Hz,

1 H, 3Ј-H), 2.27 (dd, J_6,5A= J_6,5B= 7.5 Hz, 2 H, 6-H₂), AB signal

(δ_A= 2.43, δ_B= 2.51, J_AB= 16.5 Hz, in addition split by J_A,3

8.8, J_B,3= 3.4 Hz, 2 H, 2-H₂), 2.99 (d, J_3-–OH,3= 4.0 Hz, 1 H, 3-

OH), 3.72 (s, 3 H, OCH₃), 4.01 (ddddd, J_3,2A≈ J_3,4≈ 8.3, J_3,2B

=

≈

Monohydrolysis of the Bis(4-bromobenzoate): A mixture of all of

this aforementioned residue, powdered NaHCO₃(39 mg,

460 mmol, 3.0 equiv.), and MeOH/THF (1:1, 3 mL) was heated un-

der reflux. After extraction with Et₂O (3ϫ 5 mL), the combined

organic extracts were washed with H₂O (2 mL) and brine (2 mL),

dried with Na₂SO₄, and filtered, and the solvents were evaporated.

The residue was purified by flash chromatography^[35][1.5 cm, frac-

tion volume 10 mL, petroleum ether/Et₂O 20:1Ǟ4:1 (#12)] and

yielded 39 (#18–23, 62%; 37% for the three steps from 8) as a

white solid. Recrystallization from n-hexane gave monocrystals

(m.p. 95–98 °C) that were suitable for X-ray diffraction analysis (cf.

Scheme 5).

J_3,3-OH≈ J_3,4≈ 4.0 Hz, 1 H, 3-H), AB signal (δ_A= 4.21, δ_B= 4.43,

J_AB= 11.4 Hz, in addition split by J_A,3Ј= 8.3, J_B,3Ј= 4.7 Hz, 2 H,

2Ј-H₂), 5.28 (s, 1 H, 5Ј-H) ppm. ¹³C NMR (100.6 MHz, CDCl₃/

TMS): δ = 11.56 (C-2ЈЈ), 20.21 (C-1ЈЈ), 22.46 (C-4), 34.39 (C-6),

35.70 (C-5), 41.14 (C-2), 45.48 (C-3Ј), 51.87 (OCH₃), 67.56 (C-3),

71.26 (C-2Ј), 103.89 (C-5Ј), 173.36 (C-1), 176.46 (C-6Ј), 195.24 (C-

4Ј) ppm. IR (film): ν = 3435, 2960, 2935, 2880, 1735, 1660, 1605,

˜

1460, 1440, 1405, 1365, 1205, 1170, 1010, 905, 825 cm^–1. HRMS

(EI, 70 eV): calcd. for C₁₄H₂₂O₅[M]⁺270.14672; found 270.14700

(Δ = +1.0 ppm).

Methyl 2-[(2S,6R,9R)-9-Ethyl-10-oxo-1,7-dioxaspiro[5.5]undecan-2-

yl]acetate (43) as a 90:10 mixture with Methyl 2-[(2S,6S,9R)-9-

Ethyl-10-oxo-1,7-dioxaspiro[5.5]undecan-2-yl]acetate (epi-43)

The ee of (S)-39 was determined by chiral HPLC: Chiralcel AD-H

column, n-heptane/EtOH (85:15), 1.0 mLmin^–1, 20 °C isothermal;

λ_detector= 245 nm; t_{r (S)-39}= 14.7 min, t_{r (R)-39}= 10.5 min [(S)-

39:(R)-39 = 96.6:3.4 (93.2% ee)]. [α]²_D⁰= –14.6 (c = 1.12 in CHCl₃,

1

10 cm). H NMR (400.1 MHz, CDCl₃/TMS): δ = 0.92 (dd, J_4,3A

J_4,3B= 7.4 Hz, 3 H, 4-H₃), AB signal (δ_A= 1.79, δ_B= 1.91, J_AB

=

13.5 Hz, in addition split by J_A,2= 9.2, J_A,4= 7.4 Hz,J_B,4= 7.5,

J_B,2= 5.9 Hz, 2 H, 3-H₂), 2.05 (tt, J_2Ј,1Ј= J_2Ј,3Ј= 7.3 Hz, 2 H, 2Ј-

H₂), 2.79–2.86 (m, 4 H, 1-H₂, 3-H₂), 3.30 (dddd, J_2,3A= 9.2, J_2,1A

= 7.3, J_2,1B= J_2,3B= 5.7 Hz, 1 H, 2-H), AB signal (δ_A= 4.34, δ_B

= 4.45, J_AB= 10.9 Hz, in addition split by J_A,2= 7.4, J_B,2= 5.7 Hz,

2 H, 1-H₂), 5.28 (s, 1 H, 5Ј-OH), 6.70 (s, 1 H, 7Ј-H),* 7.00 (s, 1 H,

4Ј-H),* AABB signal centered at δ = 7.56 (2 H, 3^Ar-H)* and δ =

7.86 (2 H, 2^Ar-H)* ppm; ¹³C NMR (100.6 MHz, CDCl₃/TMS): δ

= 12.17 (C-4), 24.21 (C-3), 25.78 (C-2Ј), 32.27 (C-3Ј),* 32.90 (C-

1Ј),* 39.81 (C-2), 69.43 (C-1), 112.04 (C-7Ј), 123.29 (C-4Ј), 124.71

(C-5Ј), 128.20 (C-1^Ar),* 129.29 (C-4^Ar),* 131.24 (C-2^Ar),** 131.81

(C-3^Ar),** 136.39 (C-8Ј),* 143.91 (C-9Ј),* 152.81 (C-6Ј),* 166.49

(CO₂Ar) ppm; * assignment corroborated by an HMBC experi-

ment; ** assignment corroborated by an HSQC experiment. IR

In-Situ Generation of Compounds 42 and 43: (S)-{[RuCl(BINAP)]₂-

(μ-Cl)₃}·NH₂Me₂(45 mg, 27 μmol, 0.5 mol-%; purchased from

ABCR) was added to a solution of β-keto ester 42 (1.46 g,

5.44 mmol, 1.0 equiv.) in degassed EtOH (50 mL) at room temp.

The mixture was stirred for 16 h in an autoclave under an atmo-

sphere of H₂(4.0 bar). H₂was replaced by N₂(1.0 bar), and the

solution was stirred for a further 72 h. The solvent was evaporated

under reduced pressure. H NMR analysis of the residue revealed

the presence of a 68:32 mixture of spiroketal 43 and β-hydroxy ester

42; spiroketal epi-43 was not detected.

(film): ν = 3440, 2960, 2845, 1700, 1590, 1430, 1400, 1275, 1175,

˜

1120, 1105, 1010, 955, 850, 755, 685 cm^–1. C₂₀H₂₁BrO₃(389.28):

calcd. C 61.71, H 5.44; found C 61.48, H 5.45. The crystallographic

data of this compound are available (ref.^[67]).

1

Methyl (3S)-6-[(3R)-3-Ethyl-4-oxo-3,4-dihydro-2H-pyran-6-yl]-3-hy-

droxyhexanoate (42):

Completion of Spiroketal Formation: A solution of the crude prod-

uct from the preceding paragraph and camphorsulfonic acid

(63 mg, 0.27 mmol, 5 mol-%) in CHCl₃(15 mL) was stirred for 16 h

at room temp. After addition of NaHCO₃(satd. aq.; 20 mL), the

mixture was extracted with CH₂Cl₂(3ϫ 15 mL). The combined

organic extracts were washed with brine (20 mL), dried with

Na₂SO₄, filtered, and evaporated under reduced pressure. The resi-

due was purified by flash chromatography^[35](5 cm, fraction vol-

ume 50 mL, cyclohexane/EtOAc, 5:1) to yield a 90:10-mixture* of

(S)-{[RuCl(BINAP)]₂(μ-Cl)₃}·NH₂Me₂(38 mg, 23 μmol, 0.5 mol-

%; purchased from ABCR) was added to a solution of β-keto ester

6916

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

43 and epi-43 as a colorless oil (#9–13, 965 mg, 66%); *this ratio

was determined from the H NMR integrals of the signals at δ =

3.63 (s, 3 H, OCH₃, 43) and 3.65 (s, 3 H, OCH₃, epi-43) ppm.

of trans-44, cis-epi-44, and cis-44 [this ratio was determined from

1

the H NMR integrals over the following resonances of the crude

product: 3.36 (dd, J_8ax,8eq= J_8ax,9= 11.5 Hz, 1 H, 8-H^ax, trans-44),

3.75 (ddd, J_8ax,8eq= 11.7, J_8ax,9= 2.4, ⁴J_8ax,10= 1.4 Hz, 1 H, 8-H^ax,

cis-epi-44), and 3.86 (ddddd, J_10,10-OH= 10.6, J_10,11ax= 3.4, J_10,11eq

43: ¹H NMR (499.6 MHz, CDCl₃/TMS): δ = 0.92 (dd, J_2ЈЈ,1ЈЈA

=

2

J_2ЈЈ,1ЈЈB= 7.5 Hz, 3 H, 2ЈЈ-H₃), 1.18 (ddq, J_{1ЈЈA,1ЈЈB}= 14.3, J_1ЈЈA,9

4

= 3.1, J_10,9= 2.6, J_10,8eq= 0.5 Hz, 1 H, 10-H, cis-44) ppm]. This

2

= J_1ЈЈA,2ЈЈ= 7.2 Hz, 1 H, 1ЈЈ-H^A), 1.24 (dddd, J_3ax,3eq= J_3ax,4ax

=

mixture was separated by flash chromatography^[35][5 cm, fraction

volume 50 mL, cyclohexane/EtOAc 4:1Ǟ3:1 (#8)Ǟ2:1

(#15)Ǟ1:1 (#30)] to give pure cis-44 as a white crystalline solid

(#12–15, 86 mg, 10%), and an 89:11 mixture of trans-44 and cis-

epi-44 as a colorless oil (#20–35, 735 mg, 84%).

13.5, J_3ax,2= 11.5, J_3ax,4eq= 4.2 Hz, 1 H, 3-H^ax), 1.43 (ddd, ²J_5ax,5eq

= J_5ax,4ax= 13.4, J_5ax,4eq= 4.6 Hz, 1 H, 5-H^ax), 1.60–1.68 (2 m, 2

H, 3-H^eq, 4-H^eq), 1.72–1.85 (m, 2 H, 1ЈЈ-H^B, 5-H^eq), 1.89 (ddddd,

²J_4ax,4eq= J_4ax,3ax= J_4ax,5ax= 13.4, J_4ax,3eq= J_4ax,5eq= 4.1 Hz, 1

H, 4-H^ax), 2.34–2.44 (m, 1 H, 9-H), overlapping with AB signal (δ_A

= 2.36, δ_B= 2.44, J_AB= 15.2 Hz, in addition split by J_A,2= 4.4,

J_B,2= 9.2 Hz, 2 H, 2Ј-H₂), overlapping with AB signal (δ_A= 2.39,

δ_B= 2.44, J_AB= 13.9 Hz, 2 H, 11-H₂), 3.63 (s, 3 H, OCH₃), 3.69

trans-44: ¹H NMR (400.1 MHz, C₆D₆): δ = 0.83 (dd, J_2ЈЈ,1ЈЈA

=

J_2ЈЈ,1ЈЈB= 7.5 Hz, 3 H, 2ЈЈ-H₃), 0.96 (dddd, ²J_3ax,3eq= J_3ax,4ax= 13.1,

2

J_3ax,2= 11.5, J_3ax,4eq= 3.9 Hz, 1 H, 3-H^ax), 1.09 (ddq, J_{1ЈЈA,1ЈЈB}

=

2

(dd, J_8ax,8eq= J_8ax,9= 11.2 Hz, 1 H, 8-H^ax), 3.98 (dd, J_8eq,8ax

=

13.7, J_1ЈЈA,9= 8.1, J_1ЈЈA,2ЈЈ= 7.5 Hz, 1 H, 1ЈЈ-H^A), 1.22 (ddd,

²J_5ax,5eq= J_5ax,4ax= 13.2, J_5ax,4eq= 4.8 Hz, 1 H, 5-H^ax), 1.22–1.42

(3 m, 3 H, 3-H^eq, 4-H^eq, and 9-H), overlapping with 1.27 (dd,

²J_11ax,11eq= 12.5, J_11ax,10= 11.1 Hz, 1 H, 11-H^ax), 1.57 (dddd,

²J_5eq,5ax= 13.0, J_5eq,4ax= 4.1, J_5eq,4eq= 2.4, ⁴J_5eq,3eq= 1.4 Hz, 1 H,

11.0, J_8eq,9= 7.3 Hz, 1 H, 8-H^eq), 4.08 (dddd, J_2,3ax= 11.5, J_2,2ЈB

= 9.2, J_2,2ЈA= 4.4, J_2,3eq= 2.2 Hz, 1 H, 2-H) ppm. ¹³C NMR

(125.6 MHz, CDCl₃/TMS): δ = 11.64 (C-2ЈЈ), 17.71 (C-1ЈЈ), 18.64

(C-4), 29.81 (C-3), 34.13 (C-5), 40.83 (C-2Ј), 50.83 (C-9), 51.58

(OCH₃), 52.35 (C-11), 64.07 (C-8), 66.99 (C-2), 100.68 (C-6),

171.81 (C-1Ј), 206.54 (C-10) ppm; a NOE was observed between δ

2

5-H^eq), 1.73 (dqd, J_{1ЈЈB,1ЈЈA}= 13.7, J_1ЈЈB,2ЈЈ= 7.7, J_1ЈЈB,9= 3.5 Hz,

1 H, 1ЈЈ-H^B), 1.92 (dd, ²J_11eq,11ax= 12.4, J_11eq,10= 4.9 Hz, 1 H, 11-

= 8-H^axand 2-H. IR (film): ν = 2950, 2875, 1735, 1720, 1440,

2

H^eq), 1.94 (ddddd, J_4ax,4eq= J_4ax,3ax= J_4ax,5ax= 13.2, J_4ax,3eq

=

˜

1385, 1290, 1260, 1240, 1195, 1180, 1080, 1030, 995, 970, 835 cm^–1.

C₁₄H₂₂O₅(270.32): calcd. C 62.20, H 8.20; found C 61.99, H 8.27.

J_4ax,5eq= 4.1 Hz, 1 H, 4-H^ax), AB signal (δ_A= 2.09, δ_B= 2.37, J_AB

= 15.0 Hz, in addition split by J_A,2= 4.0, J_B,2= 9.6 Hz, 2 H, 2Ј-

H₂), 3.57 (dd, J_8ax,8eq= J_8ax,9= 11.4 Hz, 1 H, 8-H^ax), 3.37 (s, 3

2

1

epi-43: H NMR (499.6 MHz, CDCl₃/TMS): δ = 0.92 (dd, J_2ЈЈ,1ЈЈA

2

H, OCH₃), 3.66–3.75 (m, 1 H, 10-H), 3.70 (dd, J_8eq,8ax= 11.3,

= J_2ЈЈ,1ЈЈB= 7.5 Hz, 3 H, 2ЈЈ-H₃), 2.14 (ddddd, J_9,1ЈЈA= J_9,1ЈЈB

=

J_8eq,9= 5.0 Hz, 1 H, 8-H^eq), 4.13 (dddd, J_2,3ax= 11.6, J_2,2ЈB= 9.5,

J_2,2ЈA= 4.0, J_2,3eq= 2.2 Hz, 1 H, 2-H) ppm. ¹³C NMR (100.6 MHz,

C₆D₆): δ = 11.40 (C-2ЈЈ), 18.98 (C-4), 21.36 (C-1ЈЈ), 30.62 (C-3),

34.90 (C-5), 41.28 (C-2Ј), 45.55 (C-9), 45.60 (C-11), 50.94 (OCH₃),

63.06 (C-8), 66.38 (C-2), 68.29 (C-10), 98.05 (C-6), 171.36 (C-1Ј)

ppm.

7.8, J_9,8ax= 3.3, J_9,8eq= J_9,11eq= 1.3 Hz, 1 H, 9-H), 2.30 (dd,

²J_11eq,11ax= 14.7, ⁴J_11eq,9= 1.1 Hz, 1 H, 11-H^eq), 2.49 (d, ²J_11ax,11eq

2

= 14.7 Hz, 1 H, 11-H^ax), 3.65 (s, 3 H, OCH₃), 3.76 (dd, J_8eq,8ax

=

11.6, J_8eq,9= 0.9 Hz, 1 H, 8-H^eq), 4.13 (dd, J_8ax,8eq= 11.5, J_8ax,9

= 3.5 Hz, 1 H, 8-H^ax) ppm. ¹³C NMR (125.6 MHz, CDCl₃/TMS):

δ = 11.81 (C-2ЈЈ), 18.66 (C-4), 23.81 (C-1ЈЈ), 34.24 (C-5), 40.77 (C-

2Ј), 49.48 (C-11), 51.61 (OCH₃), 52.00 (C-9), 62.66 (C-8), 67.06 (C-

2), 99.90 (C-6), 208.79 (C-10) ppm.

2

cis-epi-44: δ = 0.89 (dd, J_2ЈЈ,1ЈЈA= J_2ЈЈ,1ЈЈB= 7.4 Hz, 3 H, 2ЈЈ-H₃),

2

1.41 (dd, J_11ax,11eq= 12.8, J_11ax,10= 11.7 Hz, 1 H, 11-H^ax), 1.67

2

(ddd, J_11eq,11ax= 12.7, J_11eq,10= 5.0, J_11eq,9= 1.0 Hz, 1 H, 11-

Methyl 2-[(2S,6R,9R,10R)-9-Ethyl-10-hydroxy-1,7-dioxaspiro[5.5]-

undecan-2-yl]acetate (trans-44) in a 89:11 Mixture with Methyl 2-

[(2S,6S,9R,10S)-9-Ethyl-10-hydroxy-1,7-dioxaspiro[5.5]undecan-2-

yl]acetate (cis-epi-44); and Methyl 2-[(2S,6R,9R,10S)-9-Ethyl-10-

hydroxy-1,7-dioxaspiro[5.5]undecan-2-yl]acetate (cis-44) (separately

isolated):

H^eq), AB signal (δ_A= 2.09, δ_B= 2.37, J_AB= 15.0 Hz, in addition

2

split by J_A,2= 4.0, J_B,2= 9.6 Hz, 2 H, 2Ј-H₂), 3.64 (dd, J_8eq,8ax

=

11.6, J_8eq,9= 1.7 Hz, 1 H, 8-H^eq), 3.38 (s, 3 H, OCH₃), 3.90 (ddd,

²J_8ax,8eq= 11.6, J_8ax,9= 2.4, J_8ax,10= 1.3 Hz, 1 H, 8-H^ax), 4.13–

4

4.20 (m, 1 H, 2-H) ppm. ¹³C NMR (100.6 MHz, C₆D₆): δ = 11.52

(C-2ЈЈ), 15.12 (C-1ЈЈ), 19.02 (C-4), 30.62 (C-3), 34.99 (C-5), 41.28

(C-2Ј), 40.42(C-11), 41.24 (C-9), 50.92 (OCH₃), 60.47 (C-8), 66.29,

and 66.35 (C-2 and C-10), 97.80 (C-6), 171.40 (C-1Ј) ppm. IR

(film): ν = 3435, 2955, 2875, 1740, 1440, 1385, 1290, 1225, 1200,

˜

1180, 1045, 1025, 980, 935, 880, 835, 765, 680 cm^–1. HRMS (EI,

70 eV): calcd. for C₁₄H₂₄O₅[M]⁺272.16237; found 272.16250 (Δ =

+0.5 ppm).

cis-44: m.p. 101–103 °C. [α]²_D⁰= +108 (c = 1.15 in CHCl₃, 10 cm).

¹H NMR (400.1 MHz, CDCl₃/TMS): δ = 0.93 (dd, J_2ЈЈ,1ЈЈA

=

J_2ЈЈ,1ЈЈB= 7.4 Hz, 3 H, 2ЈЈ-H₃), 1.24–1.33 (m, 1 H, 3-H^ax), overlap-

2

ping with 1.23 (ddq, J_{1ЈЈA,1ЈЈB}= 13.8, J_1ЈЈA,9= 7.3, J_1ЈЈA,2ЈЈ

=

2

7.0 Hz, 1 H, 1ЈЈ-H^A), 1.42 (ddq, J_{1ЈЈB,1ЈЈA}= 13.5, J_1ЈЈB,9= J_1ЈЈB,2ЈЈ

At –78 °C, powdered NaBH₄(134 mg, 3.54 mmol, 1.1 equiv.) was = 7.5 Hz, 1 H, 1ЈЈ-H^B), 1.47–1.65 (4 m, 4 H, 9-H, 3-H^eq, 4-H^eq,

2

added in 10 portions at intervals of 1 min to a stirred solution of

CeCl₃·7H₂O (1.32 g, 3.54 mmol, 1.1 equiv.) and spiroketals 43 and

epi-43 (90:10 mixture; 869 mg, 3.21 mmol, 1.0 equiv.) in MeOH

(20 mL). The reaction mixture was stirred at –78 °C for a further

2 h, and at room temp. for 30 min. It was then poured into

NaHCO₃(satd. aq.; 30 mL). After extraction with EtOAc (3ϫ

30 mL), the combined organic extracts were washed with brine

(30 mL), dried with Na₂SO₄, and filtered, and the solvents were

evaporated in vacuo. The residue consisted of a 80:10:10-mixture

and 5-H^eq), overlapping with 1.46 (ddd, J_5ax,4ax= 13.5, J_5ax,5eq

=

2

13.0, J_5ax,4eq= 4.2 Hz, 1 H, 5-H^ax), 1.62 (ddd, J_11ax,11eq= 14.1,

4

2

J_11ax,10= 3.4, J_11ax,9= 0.5 Hz, 1 H, 11-H^ax), 1.90 (ddddd, J_4ax,4eq

≈ J_4ax,3ax≈ J_4ax,5ax≈ 13.4, J_4ax,3eq≈ J_4ax,5eq≈ 4.1 Hz, 1 H, 4-H^ax),

1.96 (dd, J_11eq,11ax= 14.2, J_11eq,10= 3.0 Hz, 1 H, 11-H^eq), AB

2

signal (δ_A= 2.43, δ_B= 2.49, J_AB= 14.9 Hz, in addition split by

J_A,2= 4.3, J_B,2= 9.3 Hz, 2 H, 2Ј-H₂), 3.46 (d, J_10-OH,10= 10.6 Hz,

2

4

1 H, 10-OH), 3.51 (ddd, J_8eq,8ax= 11.4, J_8eq,9= 4.8, J_8eq,10

=

0.8 Hz, 1 H, 8-H^eq), 3.62 (dd, J_8ax,8eq= J_8ax,9= 11.7 Hz, 1 H, 8-

2

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6917

L. Nachbauer, R. Brückner

= J_3,3-OH= 4.1 Hz, 1 H, 3-H), 4.29 (ddd, J_2Ј,1ЈA= 8.7, J_2Ј,2Ј-OH=

FULL PAPER

H^ax), 3.71 (s, 3 H, OCH₃), 4.06 (ddddd, J_10,10-OH= 10.6, J_10,11eq

=

4

3.4, J_10,11ax= 3.1, J_10,9= 2.6, J_10,8eq= 0.5 Hz, 1 H, 10-H), 4.13 J_2Ј,3= 2.3 Hz, 1 H, 2Ј-H) ppm; * assignment corroborated by an

(dddd, J_2,3ax= 11.5, J_2,2ЈB= 9.3, J_2,2ЈA= 4.2, J_2,3eq= 2.2 Hz, 1 H,

2-H) ppm. ¹³C NMR (100.6 MHz, CDCl₃/TMS): δ = 11.23 (C-2ЈЈ), 12.37 (C-5Ј), 18.89 (C-4Ј), 22.64 (C-4), 36.28 (C-5), 39.99 and 40.16

18.27 (C-4), 20.73 (C-1ЈЈ), 30.49 (C-3), 34.47 (C-5), 41.14 (C-2Ј), (C-4ЈЈЈ, C-5ЈЈЈ), 41.28 (C-2), 45.02 (C-1Ј), 45.33 (C-6), 47.11 (C-3Ј),

41.30 (C-11), 41.36 (C-9), 51.92 (OCH₃), 60.23 (C-8), 66.13 (C-10), 51.85 (OCH₃), 64.32 (C-1ЈЈ), 67.55 (C-3), 70.74 (C-2ЈЈЈ), 73.49 (C-

HMBC experiment. ¹³C NMR (125.6 MHz, CDCl₃/TMS): δ =

67.31 (C-2), 98.00 (C-6), 172.04 (C-1Ј) ppm. IR (film): ν = 3525, 2Ј), 173.36 (C-1) ppm. IR (film): ν = 3415, 2950, 2925, 2875, 1735,

˜

2955, 2940, 2875, 1740, 1440, 1390, 1290, 1200, 1175, 1150, 1075,

1435, 1280, 1200, 1160, 1100, 1040, 920, 850, 730 cm^–1. HRMS (EI,

1050, 1030, 990, 865, 840, 805 cm^–1. C₁₄H₂₄O₅(272.34): calcd. C 70 eV): calcd. for C₁₆H₃₀O₅S₂[M]⁺366.15347; found 366.15450 (Δ

61.74, H 8.88; found C 61.29, H 8.92. The crystallographic data of

= +2.8 ppm).

compound cis-44 are available (ref.^[67]).

(S)-Methyl 3-(tert-Butyldimethylsiloxy)-6-(2-{(2R,3R)-2-(tert-butyl-

dimethylsiloxy)-3-[(tert-butyldimethylsiloxy)methyl]pentyl}-1,3-di-

thiolan-2-yl)hexanoate (46):

Isolation of Pure trans-44: At –40 °C, BF₃·Et₂O (3.3 mL, 27 mmol,

10 equiv.) was added dropwise to a solution of spiroketals trans-

44 and cis-epi-44 (89:11 mixture; 735 mg, 2.70 mmol, 1.0 equiv.) in

ethanedithiol (2.7 mL, 27 mmol, 10 equiv.). After 7 h of vigorous

stirring, NaHCO₃(satd. aq.; 20 mL) was added carefully, while stir-

ring vigorously. The mixture was allowed to reach room temp. Af-

ter extraction with Et₂O (3ϫ 10 mL), the combined organic ex-

tracts were washed with brine (10 mL), dried with Na₂SO₄, and

filtered. The Et₂O was distilled off at atmospheric pressure. Ethane-

dithiol was removed under high vacuum at room temp. The residue

was purified by flash chromatography^[35](4 cm, fraction volume

20 mL, cyclohexane/EtOAc, 2:1). The title compound (#7–16,

642 mg, 87% relative to all of the 89:11 diastereomeric mixture;

98% relative to the fraction of trans-44 in the 89:11 mixture) was

isolated as a colorless oil.

At –78 °C, TBSOTf (440 μL, 1.91 mmol, 3.5 equiv.) was added

dropwise to a solution of triol 43 (200 mg, 0.546 mmol, 1.0 equiv.)

and 2,6-lutidine (450 μL, 3.82 mmol, 7.0 equiv.) in CH₂Cl₂(6 mL).

The resulting mixture was stirred for a further 20 min. The reaction

was quenched by the addition of MeOH (3 drops). The mixture

was allowed to reach room temp. After adding H₂O (6 mL), it was

extracted with CH₂Cl₂(3ϫ 5 mL). The combined organic extracts

were washed with brine (5 mL), dried with Na₂SO₄, filtered, and

concentrated in vacuo. The residue was purified by flash

chromatography^[35][1.5 cm, fraction volume 10 mL, petroleum

ether/Et₂O 24:1Ǟ19:1 (#9)] to yield the title compound as a color-

less oil (#14–20, 320 mg, 83%). [α]²_D⁰= +11.6 (c = 0.90 in CHCl₃,

(S)-Methyl 3-Hydroxy-6-{2-[(2R,3R)-2-hydroxy-3-(hydroxymethyl)-

pentyl]-1,3-dithiolan-2-yl}hexanoate (45):

1

At –40 °C, BF₃·Et₂O (16.8 mL, 137 mmol, 60 equiv.) was added to

a well-stirred solution of spiroketal trans-44 (620 mg, 2.28 mmol,

1.0 equiv.) in ethanedithiol (23 mL, 228 mmol, 100 equiv.). After

stirring for 2 d, NaHCO₃(satd. aq.; 100 mL) was added. The mix-

ture was allowed to reach room temp. After extraction with Et₂O

(3ϫ 20 mL), the combined organic extracts were washed with brine

(10 mL), dried with Na₂SO₄, and filtered. The Et₂O was distilled

off at atmospheric pressure. Ethanedithiol was removed under high

vacuum at room temp. The residue was dissolved in MeOH

(30 mL), and combined with alumina (super I B, 15 g). After stir-

ring for 1.5 h, the solvent was evaporated under reduced pressure.

The residue was packed onto a column filled with silica gel. Flash

chromatography^[35][4 cm, fraction volume 20 mL, cyclohexane/

EtOAc 1:1Ǟ1:2 (#11)Ǟ2:1 (#27)ǞEtOAc (#33)ǞEtOAc/

MeOH, 3:1 (#38)] gave the title compound as a colorless oil (#11–

40, 678 mg, 81%). [α]²_D⁰= +6.5 (c = 1.11 in CHCl₃, 10 cm). ¹H

10 cm). H NMR (499.6 MHz, C₆D₆): δ = 0.110, 0.112, 0.12, 0.17,

0.24, and 0.27 [6 s, 6ϫ 3 H, 3 Si(CH₃)₂], 1.00, 1.02, and 1.04 [3 s,

3ϫ 9 H, 3 SiC(CH₃)₃], 1.06 (dd, J_5Ј,4ЈA= J_5Ј,4ЈB= 7.6 Hz, 3 H, 5Ј-

2

H₃), 1.48 (ddq, J_4ЈA,4ЈB= 14.9, J_4ЈA,3Ј= 7.4, J_4Ј,5Ј= 7.4 Hz, 1 H,

4Ј-H^A), 1.53–1.66 (m, 3 H, 4Ј-H^B, 4-H₂), 1.73–1.86 (m, 2 H, 5-H₂),

1.89 (ddddd, J_3Ј,1ЈЈB= J_3Ј,4ЈA= 7.5, J_3Ј,1ЈЈA= J_3Ј,4ЈB= 5.1, J_3Ј,2Ј

=

2

2.3 Hz, 1 H, 3Ј-H), 2.02–2.14 (m, 2 H, 6-H₂), 2.30 (dd, J_1ЈA,1ЈB

14.9, J_1ЈA,2Ј= 5.1 Hz, 1 H, 1Ј-H^A), AB signal (δ_A= 2.39, δ_B= 2.51,

J_AB= 14.7 Hz, in addition split by J_A,3= 5.0, J_B,3= 7.4 Hz, 2 H,

2-H₂), 2.51 (dd, ²J_1ЈB,1ЈA= 14.9, J_1ЈB,2Ј= 5.9 Hz, 1 H, 1Ј-H^B), 2.75–

2.89 (m, 4 H, 4ЈЈЈ-H₂, 5ЈЈЈ-H₂), 3.37 (s, OCH₃), AB signal (δ_A

=

3.72, δ_B= 3.77, J_AB= 9.7 Hz, in addition split by J_A,3Ј= 5.5, J_B,3Ј

= 7.1 Hz, 2 H, 1ЈЈ-H₂), 4.28 (dddd, J_3,2B= 7.3, J_3,2A= J_3,4A= J_3,4B

= 5.6 Hz, 1 H, 3-H), 4.44 (ddd, J_2Ј,1ЈA= J_2Ј,1ЈB= 5.5, J_2Ј,3Ј= 2.0 Hz,

1 H, 2Ј-H) ppm. ¹³C NMR (125.6 MHz, C₆D₆): δ = –5.25, –5.24,

–4.55, –4.48, –3.98, and –3.37 [3 Si(CH₃)₂], 13.07 (C-5Ј), 18.26,

18.45, and 18.48 [3 SiC(CH₃)₃], 19.83 (C-4Ј), 23.20 (C-5), 26.08,

26.20, and 26.35 [3 SiC(CH₃)₃], 38.13 (C-4), 39.49 and 39.75 (C-

4ЈЈЈ, C-5ЈЈЈ), 42.82 (C-2), 44.53 (C-6), 47.54 (C-1Ј), 49.44 (C-3Ј),

50.97 (OCH₃), 62.53 (C-1ЈЈ), 69.98 (C-3), 70.98 (C-2ЈЈЈ), 71.15 (C-

NMR (499.6 MHz, CDCl₃/TMS): δ = 0.97 (dd, J_5Ј,4ЈA= J_5Ј,4ЈB

=

7.5 Hz, 3 H, 5Ј-H₃), 1.25–1.41 (m, 2 H, 4Ј-H₂), 1.43–1.59 (m, 2 H,

5-H₂), 1.61–1.69 (m, 1 H, 3Ј-H), overlapping with 1.59–1.74 (m, 2

H, 4-H₂), AB signal (δ_A= 1.95, δ_B= 2.05, J_AB= 13.8 Hz, in ad-

dition split by J_A,5= 11.7, J_A,5= 4.5, J_6B,5= 11.9, J_6B,5= 4.5 Hz,

2Ј), 171.50 (C-1) ppm. IR (film): ν = 2955, 2930, 2855, 1745, 1470,

˜

2

1465, 1435, 1360, 1255, 1090, 1050, 1005, 940, 835, 775, 665 cm^–1.

C₃₄H₇₂O₅S₂Si₃(709.32): calcd. C 57.57, H 10.23, S 9.04; found C

57.65, H 10.17, S 8.97.

2 H, 6-H₂) overlapping with 1.99 (br. d, J_1ЈA,1ЈB= 14.7, J_1ЈA,2Ј

Ͻ

0.5 Hz, 1 H, 1Ј-H^A), 2.18 (dd, J_1ЈB,1ЈA= 15.1, J_1ЈB,2Ј= 8.8 Hz, 1

H, 1Ј-H^B), AB signal (δ_A= 2.44, δ_B= 2.51, J_AB= 16.3 Hz, in

addition split by J_A,3= 8.8, J_B,3= 3.4 Hz, 2 H, 2-H₂), 3.08–3.12 (2

m, 2 H, 3-OH, 1ЈЈ-OH),* 3.28–3.36 (m, 4 H, 4ЈЈЈ-H₂, 5ЈЈЈ-H₂), 3.77

2

2-[(S)-4-(tert-Butyldimethylsiloxy)-5-(1,3-dithian-2-yl)pentyl]-2-

{(2R,3R)-2-(tert-butyldimethylsiloxy)-3-[(tert-butyldimethylsiloxy)-

methyl]pentyl}-1,3-dithiolane (47):

(s, 3 H, OCH₃), 3.72–3.76 (m, 2 H, 1ЈЈ-H₂), 3.91 (d, J_2Ј-OH,2Ј

=

2.2 Hz, 1 H, 2Ј-OH),* 4.05 (ddddd, J_3,2A= J_3,4= 8.1, J_3,2B= J_3,4

6918

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

phosphate buffer (pH = 7.1 Hz, 1 m, 3 mL) and potassium sodium

tartrate (satd. aq., 10 mL). After stirring vigorously for 1.5 h with-

out cooling, the mixture was extracted with CH₂Cl₂(3ϫ 5 mL).

The combined organic extracts were washed with brine (5 mL),

dried with Na₂SO₄, filtered, and concentrated under reduced pres-

sure. The oily residue (48, 239 mg, 90%) was not purified, but was

used directly to synthesize dithiane 47. ¹H NMR (300.1 MHz,

CDCl₃/TMS): δ = 0.04 (double intensity), 0.06, 0.08, 0.09, and 0.10

[5 s, 6ϫ 3 H, 3 Si(CH₃)₂], 0.88, 0.88, and 0.89 [3 s, 3ϫ 9 H, 3

SiC(CH₃)₃], 0.95 (dd, J_5Ј,4ЈA= J_5Ј,4ЈB= 7.5 Hz, 3 H, 5Ј-H₃), 1.23–

1.41 and 1.48–1.70 (m, 7 H, 4-H₂, 5-H₂, 3Ј-H, 4Ј-H₂), 1.86–1.96

(m, 2 H, 6-H₂), 2.52 (dd, J_2,3= 5.7, J_2,1= 2.5 Hz, 2 H, 2-H₂), 3.20–

3.30 (m, 4 H, 4ЈЈЈ-H₂, 5ЈЈЈ-H₂), AB signal (δ_A= 3.54, δ_B= 3.59,

J_AB= 9.7 Hz, in addition split by J_A,3Ј= 5.8, J_B,3Ј= 6.7 Hz, 2 H,

1ЈЈ-H₂), 4.10–4.23 (m, 2 H, 2Ј-H, 3-H), 9.81 (t, J_1,2= 2.5 Hz, 1 H,

1-H) ppm.

LiClO₄(28 mg, 260 μmol, 3.0 equiv.) was dried under vacuum

(0.2 mbar) for 6 h at room temp. It was then added in one portion

to a solution of aldehyde 48 (59 mg, 86.8 μmol, 1.0 equiv.) and 1,3-

propanedithiol (30 μL, 304 μmol, 3.5 equiv.) in Et₂O (1 mL). The

resulting suspension was stirred for 3 d. CuSO₄(1 m, aq., pH ad-

justed to 5 by adding a few drops of satd. aq. NaHCO₃; 2 mL) was

added with stirring. The supernatant liquid was decanted from a

solid, which we assumed was a complex formed from Cu^IIand 1,3-

propanedithiol. The solution was extracted with petroleum ether

(4ϫ 1.5 mL). The combined organic extracts were washed with

CuSO₄(1 m aq., pH adjusted to 5 by adding a few drops of satd.

aq. NaHCO₃, 2 mL) and with brine (2 mL), dried with Na₂SO₄,

filtered, and concentrated in vacuo. The residue was purified by

flash chromatography^[35][1.5 cm, fraction volume 10 mL, petro-

leum ether/Et₂O 100:1Ǟ75:1 (#13)Ǟ60:1 (#25)] to yield the title

compound as a colorless wax (#29–39, 39 mg, 63%). [α]²_D⁰= +1.31

(c = 0.933 in CHCl₃, 10 cm). ¹H NMR (400.1 MHz, CDCl₃/TMS):

δ = 0.03, 0.04, 0.08 (double intensity), 0.10, and 0.11 [5 s, 6ϫ 3 H,

3 Si(CH₃)₂], 0.88, 0.89, and 0.90 [3 s, 3ϫ 9 H, 3 SiC(CH₃)₃], 0.95

(dd, J_{5ЈЈЈ,4ЈЈЈA}= J_{5ЈЈЈ,4ЈЈЈB}= 7.5 Hz, 3 H, 5ЈЈЈ-H₃), 1.24–1.40 (m, 2

H, 4ЈЈЈ-H₂), 1.40–1.50 (m, 2 H, 3Ј-H₂), 1.50–1.61 (m, 2 H, 2Ј-H₂),

1-Butoxy-1-(6-methoxyindan-5-yl)butan-2-one (52):

At –30 °C, iPrMgCl (2.0 m in THF, 400 μL, 781 μmol, 1.4 equiv.)

was added dropwise to a solution of aryl iodide 23 (214 mg,

781 μmol, 1.4 equiv.) in THF (2 mL). The resulting mixture was

stirred for 1.5 h, and then it was added dropwise to a solution of

triflate 7b (163 mg, 558 μmol, 1.0 equiv.) and ZnCl₂(11 mg,

83.7 μmol, 15 mol-%) in THF (2 mL) at 0 °C. The mixture was

stirred for a further 2.5 h. The reaction was quenched by the ad-

dition of NH₄Cl (half-saturated aq., 4 mL). The resulting mixture

was extracted with petroleum ether (3ϫ 5 mL). The combined or-

ganic extracts were dried with Na₂SO₄, filtered, and concentrated

in vacuo. The residue was purified by flash chromatography^[35]

(2 cm, fraction volume 20 mL, petroleum ether/Et₂O, 19:1) to yield

the title compound as a colorless oil (#11–14, 38 mg, 23%). [α]²_D⁰

= +67.5 (c = 0.80 in CHCl₃, 10 cm). ¹H NMR (400.1 MHz, CDCl₃/

TMS): δ = 0.89 (t, J_4Ј,3Ј= 7.4 Hz, 3 H, 4Ј-H₃), 0.99 (t, J_4,3= 7.4 Hz,

3 H, 4-H₃), 1.38 (m, 2 H, 3Ј-H), 1.59 (m, 2 H, 2Ј-H), 2.06 (tt, J_2ЈЈ,1ЈЈ

= J_2ЈЈ,3ЈЈ= 7.4 Hz, 2 H, 2ЈЈ-H₂), AB signal (δ_A= 2.44, δ_B= 2.49,

J_AB= 17.5 Hz, in addition split by J_A,4Ј= 7.4, J_B,4Ј= 7.7 Hz, 2 H,

3-H₂), 2.83 (t, J_1ЈЈ,2ЈЈ= 7.4 Hz, 2 H, 1ЈЈ-H₂), 2.88 (t, J_3ЈЈ,2ЈЈ= 7.4 Hz,

2 H, 3ЈЈ-H₂), AB signal (δ_A= 3.40, δ_B= 3.43, J_AB= 9.0 Hz, in

addition split by J_A,2Ј= J_B,2Ј= 6.8 Hz, 2 H, 1Ј-H₂), 3.81 (s, 3 H,

OCH₃), 5.15 (s, 1 H, 1-H), 6.79 (s, 1 H, 4Ј-H), 7.17 (s, 1 H, 7Ј-H)

ppm. ¹³C NMR (100.6 MHz, CDCl₃/TMS): δ = 7.71 (C-4), 13.97

(C-4Ј), 19.39 (C-3Ј), 25.73 (C-2), 31.90 (C-3), 32.07 (C-2Ј), 32.17

(C-1ЈЈ), 33.41 (C-3ЈЈ), 55.89 (OCH₃), 69.43 (C-1Ј), 81.26 (C-1),

107.39 (C-4ЈЈ), 123.50 (C-6ЈЈ), 123.90 (C-7ЈЈ), 136.34 (C-9ЈЈ), 145.77

(C-8ЈЈ), 156.18 (C-5ЈЈ), 209.45 (C-2) ppm; an HMBC experiment

(100.6 MHz/400.1 MHz, CDCl₃/TMS) showed long-range C,H

correlations: (1) of C-5 with 1-H; (2) of C-5 with 1-H; (3) of C-2

with 3-H₂; (4) of C-6Ј with 1-H; and (5) of C-1ЈЈ with 1-H. A

NOESY experiment (400.1 MHz) revealed spatial proximities (1)

between 1-H and 3-H₂; and (2) between 1-H and 1Ј-H₂.

1.65 (ddddd, J_{3ЈЈЈ,4ЈЈЈA}= 7.7, J_{3ЈЈЈ,1ЈЈЈЈB}= 6.6, J_{3ЈЈЈ,1ЈЈЈЈA}= J_{3ЈЈЈ,4ЈЈЈB}

=

5.7, J_{3ЈЈЈ,2ЈЈЈ}= 2.1 Hz, 1 H, 3ЈЈЈ-H), 1.80–1.85 (m, 2 H, 5Ј-H₂), 1.85–

2

1.94 (m, 3 H, 1-H₂, 5ЈЈ-H^ax), 2.00 (dd, J_{1ЈЈЈA,1ЈЈЈB}= 14.8, J_{1ЈЈЈA,2ЈЈЈ}

2

= 5.5 Hz, 1 H, 1ЈЈЈ-H^A), 2.11 (ddddd, J_{5ЈЈeq,5ЈЈax}= 13.9, J_{5ЈЈeq,6ЈЈeq}

= 4.8 Hz,* J_{5ЈЈeq,4ЈЈeq}= 4.2 Hz,* J_{5ЈЈeq,6ЈЈax}= 3.3, J_{5ЈЈeq,6ЈЈax}= 2.8 Hz,

2

1 H, 5ЈЈ-H^eq), 2.25 (dd, J_{1ЈЈЈB,1ЈЈЈA}= 14.9, J_{1ЈЈЈB,2ЈЈЈ}= 5.6 Hz, 1 H,

2

1ЈЈЈ-H^B), 2.65–2.75 (m, 4ЈЈ-H^eq, 2 H, 6ЈЈ-H₂), 2.89 (ddd, J_{4ЈЈax,4ЈЈeq}

= 14.0, J_{4ЈЈax,5ЈЈax}= 11.3, J_{4ЈЈax,5ЈЈeq}= 2.7 Hz, 1 H, 4ЈЈ-H^ax), 3.19–

3.29 (m, 4 H, 4-H₂and 5-H₂), AB signal (δ_A= 3.53, δ_B= 3.59, J_AB

= 9.6 Hz, in addition split by J_A,3ЈЈЈ= 5.8, J_B,3ЈЈЈ= 6.8 Hz, 2 H,

1ЈЈЈЈ-H₂), 3.95 (dddd, J_4Ј,5ЈA≈ J_4Ј,5ЈB≈ J_4Ј,3ЈA≈ J_4Ј,3ЈB≈ 6.0 Hz, 1

H, 4Ј-H), 4.10 (dd, J_2ЈЈ,5ЈA≈ J_2ЈЈ,5ЈB≈ 7.2 Hz, 1 H, 2ЈЈ-H), 4.16 (ddd,

J_{2ЈЈЈ,1ЈЈЈA}≈ J_{2ЈЈЈ,1ЈЈЈB}≈ 5.5, J_{2ЈЈЈ,3ЈЈЈ}= 2.3 Hz, 1 H, 2ЈЈЈ-H) ppm;

* values interchangeable. ¹³C NMR (100.6 MHz, CDCl₃/TMS): δ

= –5.31, –5.27, –4.39, –4.16, –4.13, and –3.58 [6 Si(CH₃)₂], 12.83

(C-5ЈЈЈ), 18.19, 18.28, and 18.30 [3 SiC(CH₃)₃], 19.62 (C-4ЈЈЈ), 22.56

(C-2Ј), 26.05, 26.05, and 26.18 [3 SiC(CH₃)₃], 30.18 and 30.69 (C-

4ЈЈ, C-6ЈЈ), 37.74 (C-3Ј), 39.46 and 39.66 (C-4, C-5), 42.82 (C-5Ј),

43.94 (C-6Ј), 44.22 (C-2ЈЈ), 46.86 (C-1ЈЈЈ), 48.99 (C-3ЈЈЈ), 62.15 (C-

1ЈЈЈЈ), 68.76 (C-4Ј), 70.72 (C-2ЈЈЈ), 70.77 (C-2) ppm. IR (film): ν =

˜

2955, 2930, 2895, 2855, 1470 1465, 1255, 1090, 1050, 935, 835, 775,

670 cm^–1. HRMS (EI, 70 eV): calcd. for C₃₂H₆₇O₃S₄Si₃[M]⁺

711.32809; found 711.32840 (Δ = +0.4 ppm).

(S)-3-(tert-Butyldimethylsiloxy)-6-(2-{(2R,3R)-2-(tert-butyldi-

methylsiloxy)-3-[(tert-butyldimethylsiloxy)methyl]pentyl}-1,3-di-

thiolan-2-yl)hexanal (48):

Supporting Information (see footnote on the first page of this arti-

cle): Experimental procedure and characterization for triflates rac-

1

At –78 °C, DiBAH (1 m in n-hexane, 470 μL, 470 μmol, 1.2 equiv.)

was added dropwise to a solution of ester 46 (279 mg, 393 μmol,

7a and (S)-7b, copies of the H and ¹³C NMR spectra for the syn-

thesized compounds, determination of the enantiopurity of the X-

1.0 equiv.) in CH₂Cl₂(5 mL). The resulting mixture was stirred for rayed crystal of (S)-39, and X-ray crystallography of compounds

a further 30 min. The reaction was quenched by the addition of aq. (S)-39 and cis-44.

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6919

L. Nachbauer, R. Brückner

FULL PAPER

Larsen, J.-C. Leblanc, A. Mortensen, D. Parent-Massin, I.

Pratt, I. M. C. M. Rietjens, I. Stankovic, P. Tobback, T. Vergui-

eva, R. A. Woutersen, EFSA Journal 2009, 7, 1412–1427

(http://ww w .efsa.europa.eu/en/efsajournal/pub/1412.htm); b)

http://ww w .codexalimentarius.net/gsfaonline/additives/de-

tails.html?id=208&lang=en.

Total syntheses of the aglycon (1b) of rimocidin (1a): a) G. K.

Packard, Y. Hu, A. Vescovi, S. D. Rychnovsky, Angew. Chem.

2004, 116, 2882–2886; Angew. Chem. Int. Ed. 2004, 43, 2822–

2826; b) A. B. Smith III, M. A. Foley, S. Dong, A. Orbin, J.

Org. Chem. 2009, 74, 5987–6001.

Acknowledgments

L. N. expresses his gratitude to Dr. M. Keller, Institut für

Organische Chemie und Biochemie, Universität Freiburg for per-

forming the X-ray diffraction measurements and numerous NMR

analyses. We thank Dr. Klaus Ditrich (BASF Corporation, Lud-

wigshafen, Germany) for a donation of enantiomerically pure α-

hydroxy ester 7b.

[14]

[15]

[1] J. W. Davisson, F. W. Tanner, A. C. Finlay, I. A. Solomons,

Antibiot. Chemother. 1951, 1, 289–290.

[2] A. C. Cope, R. K. Bly, E. P. Burrows, O. J. Ceder, E. Ciganek,

B. T. Gillis, R. F. Porter, H. E. Johnson, J. Am. Chem. Soc.

1962, 84, 2170–2181.

For the total synthesis of amphotericin B, see: a) K. C. Nico-

laou, T. K. Chakraborty, Y. Ogawa, R. A. Daines, N. S.

Simpkins, G. T. Furst, J. Am. Chem. Soc. 1988, 110, 4660–4672;

b) K. C. Nicolaou, R. A. Daines, J. Uenishi, W. S. Li, D. P. Pa-

pahatjis, T. K. Chakraborty, J. Am. Chem. Soc. 1988, 110,

4672–4685; c) K. C. Nicolaou, R. A. Daines, T. K. Chakra-

borty, J. Am. Chem. Soc. 1988, 110, 4685–4696; d) K. C. Nico-

laou, R. A. Daines, Y. Ogawa, T. K. Chakraborty, J. Am.

Chem. Soc. 1988, 110, 4696–4705; for the total synthesis of the

methyl ester of 35-deoxyamphotericin B, see: e) A. M. Szpil-

man, D. M. Cereghetti, N. R. Wurtz, J. M. Manthorpe, E. M.

Carreira, Angew. Chem. 2008, 120, 4407–4410; Angew. Chem.

Int. Ed. 2008, 47, 4335–4338; f) A. M. Szpilman, J. M. Man-

thorpe, E. M. Carreira, Angew. Chem. 2008, 120, 4411–4414;

Angew. Chem. Int. Ed. 2008, 47, 4339–4342; for the formal total

synthesis of amphotericinolide B, see: g) D. Boschelli, T. Take-

masa, Y. Nishitani, S. Masamune, Tetrahedron Lett. 1985, 26,

5239–5242; h) R. M. Kennedy, A. Abiko, S. Masamune, Tetra-

hedron Lett. 1988, 29, 447–450; i) R. M. Kennedy, A. Abiko,

T. Takemasa, H. Okumoto, S. Masamune, Tetrahedron Lett.

1988, 29, 451–454; j) for the relay synthesis of unnatural glycos-

ides of amphoteronolide B, see: M. P. Croatt, E. M. Carreira,

Org. Lett. 2011, 13, 1390–1393.

[3] R. C. Pandey, K. L. Rinehart Jr, J. Antibiot. 1977, 30, 146–157.

[4] a) L. Falkowski, J. Golik, J. Zielinski, E. Borowski, J. Antibiot.

1976, 29, 197–198; b) P. Sowin´ski, J. Pawlak, E. Borowski, P.

Gariboldi, J. Antibiot. 1995, 48, 1288–1291.

[5] The C¹²–C¹⁸portion of rimocidin (1a) is the same as the C¹⁴–

C²⁰section of amphotericin B, nystatin A₁, and candidin. It is

also the same as the C¹²–C¹⁸section of pimaricin.

[6] a) For the isolation from Streptomyces nodosus, see: J. Vander-

putte, J. L. Wachtel, E. T. Stiller, Antibiot. Ann. 1956, 587–591;

for the connectivity and configuration, see: b) W. Mechlinski,

C. P. Schaffner, P. Ganis, G. Avitabile, Tetrahedron Lett. 1970,

11, 3873–3876; c) P. Ganis, G. Avitabile, W. Mechlinski, C. P.

Schaffner, J. Am. Chem. Soc. 1971, 93, 4560–4564.

[7] For the isolation from Streptomyces noursei, see: a) E. L.

Hazen, R. Brown, Science 1950, 112, 423–430; b) E. L. Hazen,

R. Brown, Proc. Soc. Exp. Biol. Med. 1951, 76, 93–98; for the

connectivity, see: c) C. N. Chong, R. W. Rickards, Tetrahedron

Lett. 1970, 11, 5145–5149; d) E. Borowski, J. Zielin´ski, L. Fal-

kowski, T. Zimin´ski, J. Golik, P. Kołodziejczyk, E. Jereczek,

M. Gdulewicz, Yu. Shenin, T. Kotienko, Tetrahedron Lett.

1971, 12, 685–690; e) for the relative configuration, see: J.-M.

Lancelin, F. Paquet, J.-M. Beau, Tetrahedron Lett. 1988, 29,

2827–2830; for the absolute configuration, see: f) K. C. Nico-

laou, K. H. Ahn, Tetrahedron Lett. 1989, 30, 1217–1220; g) J.-

M. Lancelin, J.-M. Beau, Tetrahedron Lett. 1989, 30, 4521–

4524.

[8] a) For the isolation from Streptomyces viridoflavus, see: W. A.

Taber, L. C. Vining, S. A. Waksman, Antibiot. Chemother.

1954, 4, 455–461; b) for the connectivity, see: E. Borowski, L.

Falkowski, J. Golik, J. Zielin´ski, T. Zimin´ski, W. Mechlin´ski,

E. Jereczek, P. Kołodziejczyk, H. Adlercreutz, C. P. Schaffner,

S. Neelakantan, Tetrahedron Lett. 1971, 12, 1987–1992; c) for

the absolute configuration, see: J. Pawlak, P. Sowinski, E.

Borowski, P. Gariboldi, J. Antibiot. 1993, 46, 1598–1603.

[9] a) For the isolation from Streptomyces natalensis, see: A. P.

Struyk, I. Hoette, G. Drost, J. M. Waisvisz, T. van Eek, J. C.

Hoogerheide, Antibiot. Ann. 1957–1958, 5, 878–885; b) for the

connectivity, see: B. T. Golding, R. W. Richards, W. E. Meyer,

J. B. Patrick, M. Barber, Tetrahedron Lett. 1966, 7, 3551–3557;

c) for the absolute configuration, see: J.-M. Lancelin, J.-M.

Beau, J. Am. Chem. Soc. 1990, 112, 4060–4061.

[10] The compound known as “pimaricin” was also described as

natamycin (a) H. Brik, Analytical Profiles of Drug Substances

1981, 10, 513–561) and tennecetin (b) P. V. Divekar, J. L.

Bloomer, J. F. Eastham, D. F. Holtman, D. A. Shirley, Antibiot.

Chemother. 1961, 11, 377–380).

[11] Reviews on the biological activity of polyol-polyene antibiotics:

a) J. Bolard, Biochim. Biophys. Acta 1986, 864, 257–304; b)

S. C. Hartsel, C. Hatch, W. Ayenew, J. Liposome Res. 1993, 3,

377–408; c) ref.^[12]

[12] Review on the use of polyol-polyene antibiotics in human ther-

apy: S. B. Zotchev, Curr. Med. Chem. 2003, 10, 211–223.

[13] Pimaricine is an important food preservative (E235): a) F. Agu-

ilar, U. R. Charrondiere, B. Dusemund, P. Galtier, J. Gilbert,

D. M. Gott, S. Grilli, R. Guertler, J. Koenig, C. Lambré, J.-C.

1

For the synthesis of the C –C¹⁹polyol fragment of nystatin A₁,

[16]

see a) I. Kadota, Y. Hu, G. K. Packard, S. D. Rychnovsky,

Proc. Natl. Acad. Sci. USA 2004, 101, 11992–11995; for the

synthesis of the C¹–C¹³fragment of nystatin A₁, see: b) : G.

Solladié, N. Wilb, C. Bauder, J. Org. Chem. 1999, 64, 5447–

5452.

1

19

[17]

[18]

[19]

For the synthesis of the C –C polyol fragment of candidin

see ref.^[16a]

.

For the total synthesis of pimaricin, see: A. J. Duplantier, S.

Masamune, J. Am. Chem. Soc. 1990, 112, 7079–7081.

a) L. Nachbauer, R. Brückner, Eur. J. Org. Chem., manuscript

to be submitted; b) R. Kramer, R. Brückner, Eur. J. Org.

Chem., manuscript to be submitted; c) S. Kamptmann, R.

Brückner, Eur. J. Org. Chem., manuscript to be submitted; d)

R. Brückner, Tetrahedron Lett. 1988, 29, 5747–5750.

A 1,2-diol that represents a building-block analogous to epox-

ide 3 was prepared by S. Hanessian, S. P. Sahoo, M. Botta,

Tetrahedron Lett. 1987, 28, 1147–1150.

A. Parenty, X. Moreau, J.-M. Campagne, Chem. Rev. 2006,

106, 911–939, and references cited therein.

For a review, see: B.-T. Gröbel, D. Seebach, Synthesis 1977,

357–402.

For a review, see: P. W. Rabideau, Z. Marcinow, Org. React.

1992, 42, 1–334.

[20]

[21]

[22]

[23]

[24]

[25]

For a review, see: G. Hilt, D. F. Weske, Chem. Soc. Rev. 2009,

38, 3082–3091.

Twice the usual number of functional groups also result when

the substrate of a Birch reduction/ozonolysis sequence contains

two aromatic rings. This was beautifully exploited by: Z. Wang,

D. Deschênes, J. Am. Chem. Soc. 1992, 114, 1090–1091.

A. J. Birch, P. Fitton, D. C. C. Smith, D. E. Steere, A. R. Stel-

fox, J. Chem. Soc. 1963, 2209–2216.

K. Tatsuta, Tetrahedron Lett. 1993, 34, 4957–4960.

D. A. Evans, J. A. Gauchet-Prunet, E. M. Carreira, A. B. Cha-

rette, J. Org. Chem. 1991, 56, 741–750.

[26]

[27]

[28]

[29]

C. L. Kirkemo, J. D. White, J. Org. Chem. 1985, 50, 1316–1319.

6920

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

[30] This reagent/solvent combination was described by: W. H. Per-

kin, R. Robinson, J. Chem. Soc. Trans. 1911, 99, 775–792.

[31] Demethylation either prior to the acylation (i.e., effecting the

reversal 22Ǟ21) or after the acylation (i.e., causing 24Ǟ50)

was not acceptable. This is because we were unable to reintro-

duce a methyl group into compound 50: the phenolate func-

tionality, which would be required for this, was alkylated faster

intramolecularly (Ǟ51) than intermolecularly (Ǟ24):

nances of the following protons: δ = 3.84 (s, 3H, OCH₃, 23)

and 3.78 (s, 3H, OCH₃, 22) ppm.

[45] For the iodine/Mg exchange in 1-alkoxy-2-iodobenzenes with

iPrMgCl, see: F. F. Kneisel, Y. Monguchi, K. M. Knapp, H.

Zipse, P. Knochel, Tetrahedron Lett. 2002, 43, 4875–4879.

[46] A reaction between triflate 7b and Grignard reagent 6a oc-

curred nevertheless. It provided 61% of a product whose ¹H

NMR spectrum indicated an indane scaffold and a butoxy

chain. We assign this compound to be α-butoxy ketone 52. We

interpret this transformation as being the result of an initial

attack of the Grignard reagent at the C=O bond of the ester

rather than at the C–O bond of the triflate. Subsequent ring-

closure leads to an epoxide (i.e., 53). The latter undergoes a

semipinacol rearrangement,^[47,48]which creates an ether and a

carbonyl group simultaneously.

[32] AlCl₃is known to demethylate ortho-methoxyacetophenones

(like 24) selectively while leaving constitutionally different

methoxy groups untouched, see: a) K. A. Parker, J. J. Petraitis,

Tetrahedron Lett. 1981, 22, 397–400; b) T.-t. Li, Y. L. Wu, J.

Am. Chem. Soc. 1981, 103, 7007–7709; c) Y. Kawamura, H.

Takatsuki, F. Torii, T. Horie, Bull. Chem. Soc. Jpn. 1994, 67,

511–515.

[33] For the activation of 2-chloropropanoic acid by this reagent

mixture, see: N. Yonezawa, Y. Tokita, T. Hino, H. Nakamura,

R. Katakai, J. Org. Chem. 1996, 61, 3551–3552.

[34] For the activation of chloroacetic acid by this reagent mixture,

see: A. Zarei, A. R. Hajipour, L. Khazdooz, Tetrahedron Lett.

2008, 49, 6715–6719.

[35] W. C. Still, M. Kahn, A. Mitra, J. Org. Chem. 1978, 43, 2923–

2925.

[36] a) K. Hoesch, Ber. Dtsch. Chem. Ges. 1915, 48, 1122–1133; b)

J. Houben, Ber. Dtsch. Chem. Ges. 1926, 59, 2878–2891.

[37] a) For a typical Houben–Hoesch acylation, see: K. C. Gulati,

S. R. Seth, K. Venkataraman, Org. Synth. 1935, 15, 70–71; Org.

Synth. Coll. 1943, 2, 522–523; b) for an early example of a

chloroacetylation of the monomethyl ether of resorcinol by this

method, see: A. Sonn, Ber. Dtsch. Chem. Ges. 1919, 52, 923–

928; c) for kinetic studies of “acylations” of benzene with

Lewis-acid-activated or protonated nitriles, see: Y. Sato, M.

Yato, T. Ohwada, S. Saito, K. Shudo, J. Am. Chem. Soc. 1995,

117, 3037–3043.

[38] This auxiliary stemmed from d-pyroglutamic acid (as described

by: Y. Kawanami, S. Murao, T. Ohga, N. Kobayashi, Tetrahe-

dron 2003, 59, 8411–8414) rather than from d-proline.

[39] a) E. J. Corey, R. K. Bakshi, S. Shibata, J. Am. Chem. Soc.

1987, 109, 5551–5553; b) E. J. Corey, R. K. Bakshi, S. Shibata,

C.-P. Chen, V. K. Singh, J. Am. Chem. Soc. 1987, 109, 7925–

7926.

[47] For the rearrangement of epoxides to give carbonyl com-

pounds, occasionally called the “Meinwald rearrangement”,

see: a) J. Meinwald, S. S. Labana, M. S. Chadha, J. Am. Chem.

Soc. 1963, 85, 582–585; b) B. Rickborn, in: Comprehensive Or-

ganic Synthesis (Eds.: I. Fleming, B. M. Trost), Pergamon, Ox-

ford, 1991, vol. 3 (Ed.: G. Pattenden), pp. 733–775.

[48] MgBr₂-catalyzed semipinacol rearrangements of isolable 2-alk-

oxy-2-aryloxiranes (like intermediate 53 in footnote^[46]) in Et₂O

to give α-alkoxyketones have been reported, see: a) T. I. Temni-

kova, N. I. Almashi, Zh. Obshch. Khim. 1953, 23, 1498–1500;

J. Gen. Chem. USSR 1953, 23, 1567–1569 (Chem. Abstr. 1954,

48, 12025f) b) C. L. Stevens, M. L. Weiner, C. T. Lenk, J. Am.

Chem. Soc. 1954, 76, 2698–2700.

[49] For ring openings of styrene oxide (29, Subst = H) with organ-

omagnesium reagents, see: J. Deniau, E. Henry-Basch, P.

Fréon, Bull. Soc. Chim. Fr. 1969, 12, 4414–4416.

[50] H. Nakao, M. Arakawa, M. Fukushima, T. Nakamura, Chem.

Pharm. Bull. 1972, 20, 1968–1979.

[40] a) S. Itsuno, K. Ito, A. Hirao, S. Nakahama, J. Chem. Soc.

Chem. Commun. 1983, 469–470; b) S. Itsuno, K. Ito, A. Hirao,

[51] M. Fukumasa, K. Furuhashi, J. Umezawa, O. Takahashi, T.

Hirai, Tetrahedron Lett. 1991, 32, 1059–1062.

S. Nakahama, J. Org. Chem. 1984, 49, 555–557; c) S. Itsuno, [52] J.-W. Jung, J. Jang, Y.-G. Suh, S.-Y. Seo, J.-K. Jung, Tetrahedron

M. Nakano, K. Miyazaki, H. Masuda, K. Ito, A. Hirao, S.

Nakahama, J. Chem. Soc. Perkin Trans. 1 1985, 2039–2044; d)

S. Itsuno, M. Nakano, K. Ito, A. Hirao, M. Owa, N. Kanda,

S. Nakahama, J. Chem. Soc. Perkin Trans. 1 1985, 2615–2619.

[41] Reviews: a) E. J. Corey, C. J. Helal, Angew. Chem. 1998, 110,

2010, 66, 6826–6831.

[53] C. Schneider, J. Brauner, Eur. J. Org. Chem. 2001, 4445–4450.

[54] G. Boireau, D. Abenhaim, E. Henry-Basch, Tetrahedron 1980,

36, 3061–3070.

[55] S. H. Bertz, G. Dabbagh, Tetrahedron 1989, 45, 425–434.

2092–2118; Angew. Chem. Int. Ed. 1998, 37, 1986–2012; b) B. T. [56] R.-D. Acker, Tetrahedron Lett. 1977, 18, 3407–3410.

Cho, Tetrahedron 2006, 62, 7621–7643.

[57] B. H. Lipshutz, J. Kozlowski, R. S. Wilhelm, J. Am. Chem. Soc.

1982, 104, 2305–2307.

[58] T. Kauffmann, C. Neiteler, S. Robbe, Chem. Ber. 1992, 125,

2409–2418.

[42] The ee of (R)-25 was determined by chiral HPLC: Chiralcel

OD-H column, n-heptane/EtOH (97:3), 0.8 mLmin^–1, 20 °C

isothermal; λ_detector= 285 nm; t_{r (R)-25}= 12.6 min, t_{r (S)-25}

11.6 min [(R)-25: (S)-25 = 99.8:0.2 (99.6% ee)].

=

[59] M. T. Didiuk, J. P. Morken, A. H. Hoveyda, Tetrahedron 1998,

[43] C. Studte, B. Breit, Angew. Chem. 2008, 120, 5531–5535; An-

gew. Chem. Int. Ed. 2008, 47, 5451–5455.

54, 117–1130.

[60] Hex₂CuMgBr ring-opens styrene oxide with complete inver-

sion of the benzylic stereocenter.^[59]The same steric course has

been reported for analogous ring-openings with allyl₂MgCuBr

and isopropenyl₂MgCuBr, yet no enantiomeric purities were

published; see: Y.-L. Chen, D. Hoppe, Tetrahedron: Asymmetry

2009, 20, 1561–1567.

[44] This value refers to the 86:14 ratio of 25 (i.e., the protonation

product of indane-based Grignard reagent 6a) vs. iodine-con-

taining indane 23 (i.e., the precursor of indane-based Grignard

reagent 6a) in a sample that we analyzed by ¹H NMR spec-

troscopy; this sample was obtained by hydrolyzing a small

amount of the solution obtained in step (b) of Scheme 4. Our

analysis consisted of determining the integral ratio of the reso-

[61] For the introduction of “higher-order cuprates”, see: a) B. H.

Lipshutz, R. S. Wilhelm, D. M. Floyd, J. Am. Chem. Soc. 1981,

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6921

L. Nachbauer, R. Brückner

FULL PAPER

103, 7672–7674. For mechanistic and structural aspects of

“higher-order cuprates”, see: b) B. H. Lipshutz, S. Sharma,

E. L. Ellsworth, J. Am. Chem. Soc. 1990, 112, 4032–4034; c) E.

Nakamura, N. Yoshikai, Bull. Chem. Soc. Jpn. 2004, 77, 1–12;

d) R. M. Gschwind, Chem. Rev. 2008, 108, 3029–3053; e) N.

Yoshikai, E. Nakamura, Chem. Rev. 2012, 112, 2339–2372.

a) H. Imogaï, M. Larchevêque, Tetrahedron: Asymmetry 1997,

8, 965–972; b) C. Botuha, M. Haddad, M. Larchevêque, Tetra-

hedron: Asymmetry 1998, 9, 1929–1932.

The ee of (S)-8 was determined by chiral HPLC: Chiralcel OD-

H column, n-heptane/iPrOH (200:1), 0.8 mLmin^–1, 15 °C iso-

thermal; λ_detector= 230 nm; t_{r (S)-8}= 21.2 min, t_{r (R)-8}= 23.6 min

[for entry 13 in Table 2: (S)-8:(R)-8 = 97.6:2.4 (95.1% ee)].

M. Miyazawa, N. Ishibashi, S. Ohnuma, M. Miyashita, Tetra-

hedron Lett. 1997, 38, 3419–3422.

For the ring-opening of an alkyl-substituted oxirane with

MeLi, see: J. L. Coke, A. B. Richon, J. Org. Chem. 1976, 41,

3516–3517.

nized by the olefin singlet at δ = 5.00 ppm and enol singlet at

δ = 12.03 ppm.

[75] In Scheme 7, the positional numbers of spiroketal ketones 43

and epi-43 use the numbering of rimocidin (1a). In contrast,

the Exp. Section features positional numbers in accordance

with the IUPAC nomenclature.

[76] According to L. DiMichele, S. A. King, A. W. Douglas, Tetra-

[62]

[63]

+

hedron: Asymmetry 2003, 14, 3427–3430, Et₂NH₂[RuCl(S)-

–

(BINAP)]₂(μ-Cl)₃is the structure of a catalyst commonly re-

ferred to as “[RuCl₂(S)-(BINAP)]₂·NEt₃”. Accordingly, we

+

–

write Me₂NH₂[RuCl(S)-(BINAP)]₂(μ-Cl)₃.

[77] An S configuration was expected for the newly formed ste-

reocenter in β-hydroxy ester 42, since (S)-BINAP-based Noyori

hydrogenations of β-keto esters always deliver this configura-

tion.^[70]

[64]

[65]

[78] Usually Noyori hydrogenations proceed with 90% yield or

more.^[70]

[79] We suppose that pyranone-containing β-hydroxy ester 42 cy-

clizes to give spiroketals 43 and epi-43 by an HCl-catalyzed

intramolecular Michael addition. HCl is a stoichiometric by-

product of the formation of active Ru^IIcomplex 55 from its

precursor 54, see ref.^[70b]Accordingly, 0.5 mol-% 54 in the reac-

tion mixture should have given 1 mol-% 55 + 1 mol-% HCl.

[66]

Compound 39 formed triclinic crystals. Starting from the non-

¯

centrosymmetric (or acentric) space group P1, the structure of

39 could be accomodated assuming that there are two dif-

ferently orientated yet identically configured molecules per unit

cell. Combined with the anomalous dispersion^[68]of the diffrac-

tion measurement, the absolute configuration of 39 was in-

ferred to be (S). Alternatively, we could accommodate the X-

ray data of 39 – albeit not quite as well – assuming that it

¯

crystallized in the centrosymmetric space group P1. This would

have implied that 39 formed a true racemate, i.e., that 39 was

racemic. We discarded that interpretation because the X-rayed

crystals revealed a 93.2% ee when we subjected them to chiral

HPLC (for more details, see Supporting Information).

CCDC-883615 (for 39) and -883616 (for cis-44) contain the

crystallographic data for this paper. These data can be obtained

free of charge from the Cambridge Crystallographic Data Cen-

tre via www.ccdc.cam.ac.uk/data_request/cif.

The presence of a heavy atom – like bromine in 39 – gives

rise to an anomalous X-ray diffraction, and thereby allows the

determination of the absolute configuration of an optically

active sample, see: J. M. Bijvoet, A. F. Peerdeman, A. J.

van Bommel, Nature 1951, 168, 271–272.

[80] A Luche reduction of pyranone-containing β-hydroxy ester 42

delivered a 80:20 mixture of epimeric but unassigned alcohols

56 and epi-56:

[67]

[68]

[69]

[70]

For the demethylation of anisoles with MeMgI, see: R. Mech-

oulam, Y. Gaoni, J. Am. Chem. Soc. 1965, 87, 3273–3275.

a) For initial publications, see: R. Noyori, T. Ohkuma, M. Kit-

amura, H. Takaya, N. Sayo, H. Kumobayashi, S. Akutagawa,

J. Am. Chem. Soc. 1987, 109, 5856–5858; b) M. Kitamura, T.

Ohkuma, S. Inoue, N. Sayo, H. Kumobayashi, S. Akutagawa,

T. Ohta, H. Takaya, R. Noyori, J. Am. Chem. Soc. 1988, 110,

629–631; c) R. Noyori, Science 1990, 248, 1194–1199. Pertinent

reviews: d) D. J. Ager, S. A. Laneman, Tetrahedron: Asymmetry

1997, 8, 3327–3355; e) T. Ohkuma, R. Noyori, in: Comprehen-

sive Asymmetric Catalysis I (Eds.: E. N. Jacobsen, A. Pfaltz, H.

Yamamoto), Springer, Berlin, 1999, pp. 199–245; f) T.

Ohkuma, M. Kitamura, R. Noyori, in Catalytic Asymmetric

Synthesis (Ed.: I. Ojima), 2^nded., Wiley-VCH, Weinheim, Ger-

many, 2000, pp. 1–111; g) R. Noyori, Angew. Chem. 2002, 114,

2108–2123; Angew. Chem. Int. Ed. 2002, 41, 2008–2022.

Solvated electrons from Li tend to give higher yields than those

from Na, see: A. L. Wilds, S. A. Nelson, J. Am. Chem. Soc.

1953, 75, 5360–5365.

For the earliest report – to the best of our knowledge – in

which tert-butyl alcohol is used as a co-solvent for Birch re-

ductions, see: H. I. Drydes Jr, G. M. Webber, R. R. Burtner,

J. A. Cella, J. Org. Chem. 1961, 26, 3237–3245.

These dyes are O₃-sensitive, and so are capable of scavenging

O₃when it is faced with less reactive C=C bonds. When an-

other C=C bond is reactive enough to react with O₃faster than

the dye does, selective ozonolyses are feasible: T. Veysoglu,

L. A. Mitscher, J. K. Swayze, Synthesis 1980, 807–810.

The enol tautomer of dihydropyranone-containing β-keto ester

anhydro-4 was admixed in a 9:91 ratio in CDCl₃. It was recog-

This was judged from the ¹H NMR spectrum (300 MHz,

CDCl₃) of the product mixture, specifically by determining the

integral ratio of the resonances appearing at δ = 4.81 (d, J_4,3

= 5.7 Hz, 1 H, 4-H, 56) and 4.67 (d, J_4,3= 4.7 Hz, 1 H, 4-

H, epi-56) ppm. Even under neutral conditions, this mixture

decomposed within a few hours to form what we assume to be

1

spiroketal 57. While the H NMR signals (300 MHz, CDCl₃)

of the decomposition product(s) were severely overlapping, the

olefin resonances δ = 5.58 (dd, J_3,4= 10.0, J_3,2= 2.6 Hz, 1 H,

3-H) and 5.81 (d, J_4,3= 10.1 Hz, 1 H, 4-H) ppm appeared to

be assignable as indicated.

[81] For ZnBr₂-induced spiroketal formation from a 1,9-dihydrox-

yalkan-5-one in CH₂Cl₂, see: M. M. Claffey, C. J. Hayes, C. H.

Heathcock, J. Org. Chem. 1999, 64, 8267–8274.

[82] For spiroketal formation from a 1,9-dihydroxyalkan-5-one in

the presence of THF/CH₃CN/HCl (12 m aq.) (10:10:1), see:

J. L. Hubbs, C. H. Heatrock, J. Am. Chem. Soc. 2003, 125,

12836–12843.

[71]

[72]

[83] ¹H NMR spectra of a 80:20-mixture of 43 and epi-43 recorded

at 300 vs. 313 vs. 333 K were identical. This observation is con-

sistent with the interpretation that epi-43 is a diastereomer of

43, and not a conformer.

[84] In Scheme 8, the positional numbers of the spiroketal alcohols

trans-44, cis-44, and cis-epi-44 use the numbering of rimocidin

(1a). In contrast, the Exp. Section features positional numbers

in accordance with IUPAC nomenclature.

[85] M. T. Crimmins, W. G. Hollis Jr., D. M. Bankaitis-Davis, Tet-

rahedron Lett. 1987, 28, 3651–3654.

[73]

[74]

[86] a) D. A. Evans, D. L. Rieger, T. K. Jones, S. W. Kaldor, J. Org.

Chem. 1990, 55, 6260–6268; b) D. A. Evans, S. W. Kaldor, T. K.

6922

www.eurjoc.org

Eur. J. Org. Chem. 2012, 6904–6923

Synthesis of a C¹–C¹¹Building Block of Rimocidin

Jones, J. Clardy, T. J. Stout, J. Am. Chem. Soc. 1990, 112, 7001–

7031; c) J. D. White, R. Hanselmann, R. W. Jackson, W. J. Por-

ter, Y. Ohba, T. Tiller, S. Wang, J. Org. Chem. 2001, 66, 5217–

5231.

[87]

a) Y.-b. Zhao, N. E. Pratt, M. J. Heeg, K. F. Albizati, J. Org.

Chem. 1993, 58, 1300–1301; b) Y.-b. Zhao, K. F. Albizati, Tet-

rahedron Lett. 1993, 34, 575–578.

[94] For the method, see: E. J. Corey, H. Cho, C. Rücker, D. H.

Hua, Tetrahedron Lett. 1981, 22, 3455–3458.

[88]

[89]

For the reduction of a 1,7-dioxaspiro[5.5]undecan-4-one analo-

[95] TBSCl and imidazole in DMF (see E. J. Corey, A. Venkates-

warlu, J. Am. Chem. Soc. 1972, 94, 6190–6191) were too unre-

active to silylate the sterically hindered OH group at C³.

[96] For example, BF₃·Et₂O was used for the cleavage of TBS to

their respective alcohols, see: a) D. R. Kelly, S. M. Roberts,

R. F. Newton, Synth. Commun. 1979, 9, 295–299; b) S. A.

King, B. Pipik, A. S. Thompson, A. DeCamp, T. R. Verhoeven,

Tetrahedron Lett. 1995, 36, 4563–4566; c) for the use of Lewis

acids for the cleavage of silyl ethers, see: A. D. Cort, Synth.

Commun. 1990, 20, 757–760.

[97] This method was described for the conversion of an aldehyde

into a 1,3-dithiane in the presence of two TBS ethers, see: F.

Eustache, P. I. Dalko, J. Cossy, J. Org. Chem. 2003, 68, 9994–

10002.

[98] Method: Y. Kamitori, M. Hojo, R. Masuda, T. Kimura, T.

Yoshida, J. Org. Chem. 1986, 51, 1427–1431.

[99] Method: D. A. Evans, L. K. Truesdale, K. G. Grimm, S. L.

Nesbitt, J. Am. Chem. Soc. 1977, 99, 5009–5017.

[100] Method: L. F. Tietze, B. Weigand, C. Wulff, Synthesis 2000,

69–71.

gous to substrate 43 using NaBH₄/CeCl₃, see ref.^[82]

The coupling constant J_3,4in spiroketal-alcohol trans-44 could

not be read from the 3-H resonance since the latter was over-

lapping the resonance of 1-H^eq. However, J_4ax,3= 11.1 Hz

could be identified from the resonance of 4-H^ax(δ_{in C6D6}

=

1.27 ppm, dd, J_gem= 12.5, J_4ax,3= 11.1 Hz), and the 12.5 Hz

splitting was atributed to J_gembecause it also appeared in the

4-H^eqresonance (δ_{in C6D6}= 1.92 ppm, dd, J_gem= 12.4, J_4eq,3

=

4.9 Hz).^[84]

[90] These J values were extracted from the hyperfine splitting of

the resonances of 1-H^axat δ = 3.90 (ddd, J_gem= 11.6, J_1ax,2

=

4

2.4, J_1ax,3= 1.3 Hz) ppm, and 1-H^eqat δ = 3.64 (dd, J_gem

11.6, J_1eq,2= 1.7 Hz) ppm. The resonance of 2-H was a dddddd

at δ = 1.29–1.36 ppm; due to overlapping resonances by 4-H^ax

3

(trans-44) and 2-H (cis-epi-44), we could not extract J_2,1ax

,

3

4

³J_2,1eq, J_2,3ax, J_2,1ЈA, J_2,1ЈB, or J_2,4eq

.

[91] These J values were extracted from the hyperfine splitting of

the resonances of 4-H^axat δ = 1.41 (ddd, J_gem= 12.8, J_4ax,3

=

11.7 Hz, 1 H) ppm, and 4-H^eqat δ = 1.68 (ddd, J_gem= 12.7,

4

J_1eq,2= 5.0, J_1eq,3= 1.0 Hz, 1 H) ppm. The resonance of 3-

[101] For the preparation and full characterization of 5-methoxyin-

dane (22), see: H. O. House, C. J. Blankley, J. Org. Chem.

1968, 33, 53–60.

H was a ddddd at δ = 4.13–4.20 ppm; due to superimposing

resonances by 9-H (trans-44) and 9-H (cis-epi-44), we could not

3

4

decipher J_3,2eq, J_3,3-OH, J_3,4ax, J_3,4eqor J_3,1ax

.

[102] For the elemental analysis of 5-iodo-6-methoxyindane (23),

see: a) C. A. Panetta, S. C. Bunce, J. Org. Chem. 1961, 26,

4859–4866. For ¹H NMR spectroscopic data for 5-iodo-6-

methoxyindane (23), see: b) A. Ali, Z. Lu, P. J. Sinclair,

Y.-H. Chen, C. J. Smith, H. Li, C. F. Thompson, U.S. Patent,

7910592, 2011. (http://appft1.uspto.gov/netacgi/nph-Parser?

Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=/netahtml/

PTO/srchnum.html&r=1&f=G&l=50&s1=20090042892.

PGNR).

[92] Procedure: R. E. Ireland, J. P. Daub, J. Org. Chem. 1983, 48,

1303–1312.

[93] It is known that spiroketals and ethane-1,2-dithiol may un-

dergo side-reactions as well as the transketalization to give a

2,2-disubstituted 1,3-dithiolane, because at “higher tempera-

tures” such mixtures are suceptible to oxidation-state changes

through formal [1,n] hydride shifts. For instance, sapogenins,

which contain substructure 58 and ethane-1,2-dithiol

(10 equiv.)/BF₃·Et₂O (6.5 equiv.) furnished tetrahydrofuran de-

rivative 59 in 71% yield, even at room temp.; see: C. Djerassi,

O. Halpern, G. R. Pettit, G. H. Thomas, J. Org. Chem. 1959,

24, 1–6.

[103] Dry HCl gas was generated by addition of concd. aq. HCl to

CaCl₂; see: F. J. Arnáiz, J. Chem. Educ. 1995, 72, 1139–1139.

Received: August 15, 2012

Published Online: November 7, 2012

Eur. J. Org. Chem. 2012, 6904–6923

www.eurjoc.org

6923

Article Doi

Converting a birch reduction product into a polyketide: Application to the synthesis of a C1-C11 building block of rimocidin

DOI: 10.1002/ejoc.201201112

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejoc.201201112

Full text of DOI:10.1002/ejoc.201201112

Products guided by the article

R&D Labs maybe for 1442440-30-3

Relevant to this article

Hot Product