Rubrenolide, total synthesis and revision of its reported stereochemical structure

Tetrahedron 60 (2004) 5237–5252

Rubrenolide, total synthesis and revision of its reported

stereochemical structure

*

L. Thijs and B. Zwanenburg

Department of Organic Chemistry, NSR Institute, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands

Received 11 July 2003; revised 19 March 2004; accepted 14 April 2004

Abstract—In this paper the synthesis of the natural product rubrenolide is presented. Due to an error in the original proposed stereochemical

structure of rubrenolide, the synthesis was not straightforward. Application of the photo-induced rearrangement of an appropriate epoxy

diazomethyl ketone gave access to the precursor lactone with an ee of 91%. Coupling of this lactone with (4S)-2,2-dimethyl-[1,3]-dioxolane-

4-carbaldehyde gave, after some additional steps, the ﬁnal product that was identical with an authentic sample of the natural product.

1. Introduction

Rubrenolide 1 is a natural product which has been isolated

from trunk wood of the Amazonian tree Nectandra rubra of

the Lauraceae family. Its chemical constitution was ﬁrst

published in 1971.¹Actually, rubrenolide 1 was isolated as a

mixture with an analog—rubrynolide 2, which only differs

in the terminal unsaturated bond in the aliphatic side chain

Figure 2.

(Fig. 1). The biosynthesis of these compounds was also

described.²In 1977 the full structure, including assignment

of the conﬁguration at the chiral centers, was reported.³The

assignment of the conﬁguration of rubrenolide was made

according to the denotation of the structure in Figure 1 and it

was named (2S,4R)-2-[(2⁰S)-2⁰,3⁰-dihydroxypropyl]-4-(dec-

9i⁰⁰-enyl)-g-lactone.^†According to this structure the sub-

stituents at C₂and C₄have a cis relationship. It is important

to note here that in assigning the R/S descriptors great care

should be exercised by the correct use of the priority rules,⁴

which often are not straightforward. In Figure 2 the priority

of the groups attached to the chiral center C₂is depicted for

the sake of clarity.

In the paper by Franca et al.³concerning the absolute

conﬁguration of rubrenolide and rubrynolide the following

phrase is puzzling: ‘The complete assignment of all proton

resonances and coupling constants associated with the

g-lactone showed rubrenolide to possess the 2,4-trans

conﬁguration. Comparison with analogous data given by

some model cis (7) and trans (8) 2,4-disubstituted

g-butyrolactones, specially synthesized for this purpose,

yielded unequivocal proof of this fact’.

Because of the very detailed NMR studies that were carried

out on rubrenolide and the model compounds mentioned,⁵it

must be assumed that an interchange of two ﬁgures (denoted

there as 7 and 8, see quote) has taken place in the ﬁnal

paper,³leading to the incorrect statement that the sub-

stituents at C₂and C₄are positioned trans.

Figure 1.

An approach to the synthesis of rubrynolide 2, reported by

Taylor et al.⁶in 1991, was regrettably not noticed by us at

that time. This synthesis was based on a reaction of an

epoxide with an aluminum enolate, which leads to lactone 3

as a trans–cis mixture in a ratio of 85:15 (Scheme 1).

Dihydroxylation of the oleﬁnic bond with OsO₄resulted in a

mixture of four racemic components, namely cis R alcohol

Keywords: Rubrenolide; Epoxy diazomethyl ketone; Lactone.

*

Corresponding author. Tel.: þ31-24-365-3159; fax: þ31-24-365-3393;

e-mail address: zwanenb@sci.kun.nl

Throughout the text the original nomenclature will be used, describing

†

rubrenolide as a g-lactone. In Section 5 the names generated by the

Beilstein Autonom version 2.1 program will be used describing a

g-lactone as dihydrofuran-2-one.

doi:10.1016/j.tet.2004.04.037

5238

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

Scheme 1.

4a, cis S-alcohol 4b, trans R-alcohol 4c, and trans S-alcohol

4d. The₀ir conﬁgurations relative to the 4R conﬁguration are

2S,4R,2₀R for 4a, 2S,4R,2⁰S for 4b, 2R,4R,2⁰R for 4c and

2R,4R,2 S for 4d. Only one of the minor products was,

according to the NMR analysis of the mixture, identical with

rubrynolide, which implies either structure 4a or 4b. As the

diols could not be separated, they were converted into their

diacetates, which could be separated by HPLC. One product

was identical with the diacetate of rubrynolide. Taylor et al.

assumed that the originally assigned R conﬁguration of C₄

constitutes an extra challenge. The ₀retrosynthesis of

rubrenolide with the proposed (2S,4R,2 S)-conﬁguration is

outlined in Scheme 2. The stereochemistry at C4 is derived

from the epoxy diazomethyl ketone and can be introduced

as required by choosing the appropriate chiral inductor

during the Sharpless epoxidation that is used in the synthesis

of the epoxy unit. The side chain can possibly be introduced

by an alkylation reaction, which may give rise to cis/trans

mixtures that hopefully can be separated.

0

was correct, that C₂has the S conﬁguration, and that the

diacetate derived from 4b was identical to rubrynolide

diacetate, which then should have the 2S,4R,2⁰S conﬁgur-

ation. This is exactly the same conﬁguration as was reported

by Franca et al. Notwithstanding, Taylor et al. entitled their

paper as ‘Synthesis of (^)-rubrynolide and a revision of its

reported stereochemistry’.

2. Results

2.1. Synthesis of rubrenolide with the proposed

(2S,4R,2⁰S)-conﬁguration

The synthesis of the required epoxy diazomethyl ketone 12

started with decane-1,10-diol 5 which was treated with

hydrobromic acid to give the bromo alcohol 6 (Scheme 3).

After protection of the alcohol as a THP ether (compound 7)

a chain elongation with the dianion of propargylic alcohol

was carried out. The alkynol 8 was reduced with LiAlH₄to

give the trans allylic alcohol 9 with high efﬁciency. In the

next step the required chirality had to be installed by

choosing the right chiral inductor in the Sharpless epoxida-

tion.¹⁰In order to obtain the 4R conﬁguration in the natural

product 1 ^D-(2) diethyl tartrate was needed for this purpose.

The (2R,3R)-epoxy alcohol 10 thus obtained needed to be

oxidized to the corresponding carboxylic acid. The one-step

oxidation procedure with RuO₄was not possible in this case

due to the vulnerability of the THP protecting group.

It seems that Taylor et al. refer to the phrase of Franca,

which is cited above in italics, stating that rubrenolide and

rubrynolide have the 2,4-trans relationship (which is

incorrect vide supra). The total synthesis of natural

rubrenolide as carried out by the Nijmegen group,⁷

unambiguously showed that the correct absolute conﬁgur-

ation of this product is 2S,4R,2⁰R.

The total synthesis of the naturally occurring functionalized

g-lactone rubrenolide 1 offers an opportunity to test the

applicability of the synthetic methodology based on epoxy

diazomethyl ketones.⁸The key reaction in this methodology

for the synthesis of g-lactones is the photo-induced

rearrangement of epoxy diazomethyl ketones to epoxy

ketene followed by a domino reaction with an alcohol to

give g-hydroxy-a,b unsaturated esters.⁹The introduction of

the side chain in rubrenolide in a stereocontrolled manner

The Swern oxidation to the aldehyde, followed by further

oxidation with sodium chlorite¹¹to the carboxylic acid was

a perfect alternative. The thus obtained carboxylic acid 11

Scheme 2.

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5239

Scheme 3. (a) HBr (48%), pet. ether 80–100 8C (73%); (b) DHP, TsOH (94%); (c) LiNH₂6 equiv., propyn-2-ol-1, 3 equiv., (99%); (d) LiAlH₄, 2 equiv.,

(quant.); (e) ^D-(2)-DET, Ti(iOPr)₄, t-BuOOH (69%); (f) Swern oxid.(quant.); (g) NaClO₂(quant.); (h) i-butyl chloroformate, Et₃N; (i) CH₂N₂(68%, calcd on

epoxy alcohol 10).

was converted into the epoxy diazomethyl ketone 12 by

reaction with iso-butyl chloroformate and triethylamine to

give the mixed anhydride and subsequent treatment with

ethereal diazomethane. The diazo ketone 12 was puriﬁed by

chromatography and obtained in an acceptable overall yield

(Scheme 3).

For the next step, the introduction of the terminal oleﬁnic

bond, o-nitrophenylselenyl cyanide was used as the

reagent.¹⁴Although this ingredient is not very pleasant,

the formation of selenide 16 and the subsequent elimination

reaction to give lactone 17 took place smoothly (Scheme 4).

Product 17 was slightly yellow due to impurities arising

from the nitrophenylselenyl reagent.

The next step in the sequence to the g-lactone 17 is the

photo-induced rearrangement of epoxy diazomethyl ketone

12 in ethanol as the solvent. This reaction was monitored by

IR. Unexpectedly, the THP protecting group was lost

completely during this reaction (Scheme 4). The product

obtained after irradiation, i.e. the g-hydroxy-a,b-unsatu-

rated ester 13, preferably is used for further reactions

without puriﬁcation.¹²Thus, product 13 was subjected to

reduction with P2–Ni¹³as the catalyst to give the dihydroxy

carboxylic ester 14. Subsequent lactonization to g-lactone

15 was accomplished by treatment with TsOH in benzene at

reﬂux temperature.

The planned alkylation step was attempted with (4R)-4-

iodomethyl-2,2-dimethyl-[1,3]-dioxolane using the Li-

enolate of lactone 17 (Scheme 5(a)). Despite the use of a

range of reaction conditions, the desired coupling could not

be realized. Apparently, this alkylating agent is very

reluctant to undergo an S_N2 reaction with an enolate

anion. An alternative candidate electrophile for the

introduction of the side chain is (4R)-2,2-dimethyl-

[1,3]dioxolane-4-carbaldehyde 18.¹⁵

The enolate of 17 reacted smoothly with aldehyde 18 to give

Scheme 4. (a) hn, 300 nm, EtOH; (b) Ni(OAc)₂, NaBH₄(57% calcd on 12); (c) TsOH, benzene (quant.); (d) o-NO₂PhSeCN, Bu₃P (93%); (e) H₂O₂, Na₂CO₃

(68%).

Scheme 5.

5240

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

Scheme 6. (a) LDA, THF, HMPA, (80%); (b) thiocarbonyl diimidazole (64%); (c) Bu₃SnH, toluene, gentle warming, (70%); (d) Bu₃SnH, toluene, reﬂux,

(64%); (e) TsOH, MeOH, (61%).

a mixture of two major products (ratio 1:1) as was deduced

from a GLC analysis. Tentatively, structures 19 and 20 were

assigned to these condensation products (Scheme 5, line b).

If the assignment of the Aldol products 19 and 20 is correct,

then it will be possible to prepare the rubrenolide type

product with the cis as well as the trans conﬁguration. For

the transformation of the Aldol products into the ﬁnal

(400 MHz). The products were identiﬁed as the cis and trans

compounds 25 and 26 (product ratio 2:3). ‘Surprisingly,

neither of these two products was the expected

rubrenolide’.^‡

This conclusion was further substantiated as follows.

Hydrogenation of the mixture of the unsaturated compounds

23 and 24 gave a single product as was shown by GLC

analysis. Indeed, catalytic hydrogenation should produce

the cis-disubstituted lactone 27 (Scheme 7). Predictably, the

terminal oleﬁn is hydrogenated as well. This saturated

compound 27 (mp 38–41 8C) should then be identical to the

dihydrorubrenolide derivative obtained by Franca et al.³by

reaction of dihydrorubrenolide with acetone.

0

products the alcohol function at C₁has to be removed. An

attractive approach would be the conversion into a thio-

imidazolide 21 and a subsequent reduction with tributyltin

hydride¹⁶(Scheme 6).

The formation of the thio-imidazolide 21 was accomplished

in THF at room temperature for 6 h, in a yield of 64%. Some

starting material (21%) was recovered and some by-

products, mainly a mixture of 23 and 24 were isolated.

Repeating the same reaction at reﬂux temperature for 3 h

gave product 21 (55%), starting material (20%) and a

mixture of 23þ24 (22%).

The spectral data however were not in agreement with those

reported³and the melting points were different as well (lit.³

mp 47–48 8C). When the acetonide unit was hydrolyzed,

product 28 with mp 94–96 8C was obtained, which,

according to its mp and NMR spectrum^§clearly was not

dihydrorubrenolide as described by Franca et al.³(mp

dihydrorubrenolide: 106–107 8C). Thus the (2S,4R,2⁰S)

conﬁguration is not in agreement with the true structure

of rubrenolide. The unambiguous conclusion is that the

proposed absolute conﬁguration of 2S,4R,2⁰S by Franca

et al.³for dihydrorubrenolide and rubrenolide is not correct.

In addition, the results also indicate that the 2R,4R,2⁰S

conﬁguration, as in trans disubstituted lactone 26, is not

dihydrorubrenolide either.

For the reductive removal of the thio-imidazolide group in

21 the method described by Rasmussen et al.¹⁶was

followed, namely reaction with tributyltin hydride in

toluene at reﬂux temperature. A mixture of the unsaturated

compounds 23 and 24 was obtained in 64% yield. These

geometrical isomers could be separated by chromatography

1

and the structures could be assigned on the basis of the H

NMR spectra.¹⁷Unexpectedly, the product 22 had not been

obtained under these conditions. However, repeating the

reaction at lower temperature, i.e. after addition of freshly

prepared tributyltin hydride, the reaction mixture was gently

heated for 30 min, resulted in the desired product 22 in 70%

yield and almost no unsaturated product was obtained.

According to GLC this product consisted of two isomers (as

expected), however, chromatographic separation on silica

gel was not possible. Hydrolysis of the product 22 by

treatment with TsOH in methanol gave the deprotected

product, which according to GLC and TLC, was not a

mixture of isomers. However, the presence of the expected

‡

Compound 25 (cis substituted lactone) is an epimer of rubrenolide at the

C2⁰position. The trans disubstituted lactone 26 shows ¹H NMR

absorptions identical with one of the trans dihydroxy-propyl lactones

described in Ref. 6. The compound is denoted as trans-alcohol 4d in

Scheme 1.

§

The late Professor Ollis, kindly provided us with a free sample of the

natural product consisting of a mixture of rubrenolide and rubrynolide.

The two products could be separated as reported.³Rubrynolide was

hydrogenated to give dihydrorubrenolide (see Section 5).

1

two isomers was demonstrated by H NMR spectroscopy

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5241

Scheme 7. (a) H2, Pd/C, quant.; (b) TsOH, MeOH, quant.

Scheme 8. (a) Propyn-2-ol-1, 3 equiv., LiNH₂6 equiv., (90%); (b) LiAlH₄, 2 equiv., (85%); (c) (D)-(2)-DET, Ti(OiPr)₄, TBHP, (77%); (d) RuCl₃·3H₂O,

NaIO₄; (e) i-BuO(CvO)Cl, Et₃N, CH₂N₂, (47%); (f) hn, 300 nm, MeOH; Ac₂O, pyridine, (50%); (g) PdCl₂, NaBH₄, H₂, (76%); (h) KOH/EtOH, reﬂux,

15 min, TsOH, benzene, reﬂux, quant.; (i) LDA, DMPU, (4S)-2,2-dimethyl-[1,3]-dioxolane-4-carbaldehyde 38, (63%); (j) MsCl, pyridine, DBU, (96%);

(k) TsOH, MeOH, (77%); (l) Pd/C, H₂, EtOH, (84%).

2.2. Synthesis of rubrenolide with the correct (2S,4R,2⁰R)

conﬁguration

purity of the lactone 32 was determined via ¹H NMR of the

acetate 31 (by the use of chiral shift reagent) and appeared to

be 91%.¹⁹

0

As mentioned above it seemed reasonable that the C₂

conﬁguration is opposite to the original assignment. There-

fore, the synthesis of dihydrorubrenolide using (4S)-2,2-

dimethyl-[1,3]-dioxolane-4-carbaldehyde 38 (the antipode

of aldehyde 18) to install the R-chirality at C₂, was

undertaken.

The introduction of the side chain was performed by a

condensation of the lactone 32 with (4S)-2,2-dimethyl-

[1,3]-dioxolane-4-carbaldehyde²⁰38. The thus obtained cis/

trans mixture of alcohol 33 was treated with mesyl chloride

to give the corresponding mesylate, which was subjected to

an elimination reaction by treatment with DBU giving

alkene 34 as a cis/trans mixture. Catalytic hydrogenation

was best performed after deprotection of the vicinal

alcohols. Deprotected diol 35 was then hydrogenated

using Pd(C) as the catalyst to produce (2S,4R,2⁰R)-dihydro-

rubrenolide 36, which, gratifyingly, was identical to the

dihydrorubrenolide prepared from an authentic sample

of rubrenolide. The conclusion can be drawn that the

absolute conﬁguration of natural rubrenolide must be

2S,4R,2⁰R.

0

This synthesis of dihydrorubrenolide (Scheme 8) follows

the general sequence for the preparation of lactones.

Reaction of decyl bromide with the di-anion of propargyl

alcohol, reduction to the trans-allylic alcohol, Sharpless

epoxidation with ^D-(2) DET as chiral inductor, RuO₄

oxidation to the glycidic acid, activation of the acid

followed by treatment with diazomethane, gave the epoxy

diazomethyl ketone 30 in good yield. Irradiation in MeOH

and subsequent acetylation gave 31, reduction of the double

bond and ring closure resulted in the desired lactone 32. The

synthesis of this lactone was reported previously as the

reduction product of the sex pheromone cis-4R-4-(1⁰-

decenyl)-g-lactone¹⁸of the Japanese beetle.^{The enantio-

The total synthesis of natural rubrenolide, having the

terminal oleﬁnic bond, was prepared from epoxy diazo-

methyl ketone 12 as the key starting material, following the

synthetic sequence shown in Scheme 3. The photo-induced

rearrangement of 12 could be conducted in this case without

the loss of the THP protecting group, thus providing

13-OTHP, which was reduced with P2-Ni/H₂to the

g-hydroxy ester which was ring closed to the lactone 37

{

Although the conﬁguration at the chiral centre of this natural product was

assigned correctly as R, the reduction product was also given the R

conﬁguration, which is violating the CIP rules, because the change in

priority has not been implemented.

5242

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

(compare synthesis of lactone 15 shown in Scheme 4).

Condensation of lactone 37 with the 4(S)-carbaldehyde 38

smoothly led to product 39 (mixture of cis/trans isomers)

(Scheme 9). The exo double bond was introduced by

elimination of the mesylate using DBU as the base (compare

Scheme 8). The thus obtained product 40 (mixture of cis and

trans) was treated with acid to remove both protective

groups to give triol 41, which on catalytic hydrogenation

(Pd/C, H₂) afforded the single product 42, as was evident

from its narrow melting range. For the ﬁnal introduction of

the terminal oleﬁnic bond the vic. diol had to be protected as

the acetonide 43. Then the primary hydroxy group was

converted into the alkene 44, following the procedure

shown in Scheme 4 using o-nitrophenylselanyl cyanide.

Finally, removal of the acetonide protecting group gave the

target product 1, whose spectral features and mp were in full

agreement with those of the natural product that was kindly

provided by (the late) Professor Ollis.

This problem can be solved by implying the Cotton effect.

The contribution of the chiral lactone ring to the optical

rotation can be estimated by subtracting the measured

rotation of a reference compound without the lactone ring.

This, for instance, could be the open form: hydroxy acid or

hydroxy ester. The o.r.d. curves of the lactone and a

reference compound can be measured and a difference curve

can be obtained by subtraction. The sign of the difference

curve can now be related to the conﬁguration at the chiral

lactone site (a positive sign means that the conﬁguration is

R). The Cotton effect measures the rotation at the ultraviolet

absorption of the lactone function (normally between 225

and 233 nm). The sign of the Cotton effect is in agreement

with the sign of the difference curve and thus can be used to

determine the contribution of the lactone ring to the optical

rotation. Rubrenolide showed a positive Cotton effect at

226 nm (fþ1700). This means that rubrenolide has the

(R)-conﬁguration at C₄.

Accepting this assignment, two possible diastereomers

remain, ₀viz. 2S,4R,2⁰R; ₀and 2R,4R,2⁰R because the

(2S,4R,2 S) and (2R,4R,2 S)-conﬁguration were already

excluded. Closer examination of the 400 MHz NMR spectra

revealed that the difference between the synthesized product

28 (Scheme 7) and dihydrorubrenolide derived from the

natural product²²is caused by the environment of the

3. Discussion

Having established ₀that rubrenolide does not have the

proposed³(2S,4R,2 S)-conﬁguration, the question arose

what would be the alternative. The original assignment of

the conﬁguration at C₄seemed quite reliable. This was

deduced by application of Hudson’s lactone rule.²¹In

simple terms this rule states that when the lactone is drawn

in the Fischer projection (Fig. 3(a)) and the lactone ring lies

to the right then the rotation is (þ). When the lactone

structure is drawn in an alternative way (Fig. 3(b)) then the

rule states that if the hydrogen lies below the plane of the

lactone ring, the compound will have a (þ) rotation. This

Hudson lactone rule is supported by numerous examples. It

also predicts the rotation and conﬁguration of the simple

4-monosubstituted g-lactones. When more stereogenic

centers are present the situation changes. When the rule is

applied to muricatacin (Fig. 3(c)) the predicted optical

rotation is not in agreement with the conﬁguration of the

chiral center of the lactone.

0

hydroxy at C₂. Franca et al.³who reported the complete

0

structure of rubrenolide, determined the conﬁguration at C₂

by applying Horeau’s method²³to the monomethyl ether of

the primary alcohol of the natural product. The free

secondary alcohol was brought into reaction with

a-phenylbutyric anhydride in the presence of pyridine.

Hydrolysis of the remaining anhydride gave the acid with a

positive rotation. This result then was related to a Fischer

projection (Fig. 4), which according to Horeau’s method

0

leads to the S-conﬁguration at the C₂chiral center. It is

important to note that Horeau and Kagan²⁴state in a

footnote in their paper of 1964 that when a polar group is

close to the secondary alcohol, other rules for the assign-

ment of L and M in the Fischer projection may be necessary.

Scheme 9. (a) LDA, HMPA, THF, 270 8C, 16 h, (26%); (b) MsCl, Et₃N, DBU, (72%); (c) TsOH, MeOH, (70%); (d) Pd/C, H₂, quant.; (e) TsOH, acetone,

(73%); (f) o-nitro-phenylselanyl cyanide, Bu₃P, (91%); (g) H₂O₂; (h) TsOH, MeOH, (60%).

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5243

Figure 3.

Figure 4.

0

This special situation is encountered in the case of

rubrenolide because of the methoxy group and leads to a

deviant priority of M and L, with the ultimate consequence

C₄and C₂is deﬁned by the starting materials, the product

mixture AþB is to be expected.

0

that the assignment of the chirality at C₂by Franca et al.³is

not correct.

4. Concluding remarks

Concerning Scheme 5 it was assumed that a mixture of cis

and trans substituted lactone products was obtained. A more

detailed representation of the possible Aldol condensation

products is depicted in Scheme 10. The (R)-aldehyde 18 is

taken as an example.

The absolute conﬁguration of the natural products rubreno-

lide and rubrynolide has been established by a fully

stereoselective total synthesis. The originally assigned

(2S,4R,2⁰S)-conﬁguration turned out to be incorrect, the

0

conﬁguration at C₂is opposite. In the original paper of

Franca et al.³dealing with the absolute stereochemical

structure of these natural products, an incorrect and

confusing statement concerning the cis/trans geometry

was made. In the synthetic strategy effective use was

made of the chemistry of epoxy diazomethyl ketones, in

particular the photo-induced rearrangement in an alcoholic

solution leading to g-hydroxy-a,b-unsaturated esters, which

conveniently can be converted to g-lactones with a designed

chirality at C₄, when optically active substrates are used.

It should be noted that the R descriptor for the chirality

changes to S in the ﬁnal product due to a change in the

priority of the substituents ﬂanking this chiral center.

Conclusive information about these structures was obtained

by comparison with related condensations reported in the

literature and by applying the Felkin–Ahn rule²⁵for

the reaction of an enolate with a carbonyl compound. The

reaction of the Li-enolate of methyl propanoate with

carbaldehyde 18 results in the formation of a mixture of

anti/syn and anti/anti product in a ratio of 3:2 (Scheme 10,

line b).²⁶Thus this condensation exhibits an exclusive anti

selectivity with respect to the dioxolane ring. The selectivity

with respect to the ester moiety is rather low as is apparent

from the anti,syn/anti,anti ratio of 3:2. The Felkin–Ahn

model for the reaction of aldehyde 18 is shown in Figure 5.^k

This picture is in agreement with the proposed mechanism

described by Jurczak et al.²⁷By calculating minimum

energy structures of aldehyde 18, attack of the nucleophile

leads to the same result (Fig. 5).²⁸Clearly, the syn attack is

disfavored ruling out structures C and D. It is reasonable to

assume that the condensation of the enolate of 17 with

aldehyde 18 also proceeds with a high selectivity with

respect to the dioxolane ring,²⁹implying the preferred

formation of the anti/syn and anti/anti products shown in

Scheme 10, line a. Taking into account that the chirality at

5. Experimental

5.1. General remarks

The solvents used for synthetic procedures and for

chromatography were puriﬁed and dried according to

conventional methods. Melting points were determined

using a Reichert thermopan microscope and are uncor-

rected. IR spectra were recorded on a Perkin–Elmer 298

spectrophotometer or on a FTIR spectrophotometer of ATI

Mattson (Genesis Series). NMR’s were measured on

1

different instruments: routinely H spectra with a Varian

EM-390, or a Bruker AC 100 (100 MHz, FT) instrument.

More detailed spectra requiring also ¹³C data were

measured on a Bruker AM-400 (400 MHz, FT) instrument.

Optical rotations were measured with a Perkin–Elmer 241

MC automatic polarimeter. Mass spectra were run on a

double focusing VG 7070E instrument. Elemental analyses

were performed using a Carlo Erba Instruments CHNS–O

1108 element analyzer. GLC’s were run using

k

The representation of the Felkin–Ahn model was misinterpreted in the

ﬁrst draft of this paper as was pointed out by one of the referees.

5244

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

Scheme 10.

Figure 5.

Hewlett–Packard 5790 or 5890 instruments, equipped with

apolar capillary columns. TLC’s were performed using

glass plates coated with 25 mm silica (Merck 60 F-254).

Spots were visualized with a 6.2% H₂SO₄solution (1 L)

containing ammonium molybdate (42 g) and cerium

ammonium sulfate (3.6 g), followed by charring. For

‘normal’ chromatography silica gel 60 (Baker) was used.

For ﬂash chromatography Kieselgel 60H (Merck) was

applied.

ether (bp 80–100 8C) during 60 h. The organic layer was

collected and shaken with solid K₂CO₃. After ﬁltration the

ﬁltrate was concentrated in vacuo and the residue was

puriﬁed by ﬂash chromatography (hexane–diethyl ether

8:1, removal of dibromide, followed by diethyl ether). Yield

of 6 (colorless oil) 47.0 g (73%). ¹H NMR (CDCl₃): d 3.60

(t, 2H, CH₂OH), 3.35 (t, 2H, CH₂Br), 1.40 (m, 16H) ppm.

5.1.2. 2-(10-Bromo-decyloxy)-tetrahydropyran 7. To a

cooled (0 8C) solution of 10-bromo-decan-1-ol 6 (27.5 g

116 mmol) in dichloromethane (100 mL) a small amount of

p-TsOH was added, followed by dropwise addition of

dihydropyran (10 g, 119 mmol). The reaction mixture was

stirred for 2 h, diluted with diethyl ether and washed with

5.1.1. 10-Bromo-decan-1-ol 6.³⁰Decane-1,10-diol

5

(47.0 g, 0.27 mol) was mixed with hydrobromic acid

(380 mL, 47%) in a Dean–Stark apparatus (1 L size) and

the mixture was extracted continuously with petroleum

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5245

aqueous NaHCO₃solution. The organic phase was dried

(MgSO₄), ﬁltered and concentrated. The residue was

puriﬁed by ﬂash chromatography (hexane–ether 6:1) giving

used as such in the next step. A small amount was puriﬁed

by chromatography for further analysis. IR (neat): n_max

3360 (OH), 1662 (CvC) cm²¹. NMR (400 MHz) (CDCl₃):

d 5.73–560 (2H, m, HCvCH), 4.57 (1H, m, O–CH–O),

4.07 (2H, m, CH₂OH), 3.86 (1H, m, CHH–O, THP ring),

3.71 (1H, d t, CHH–O, chain), 3.49 (1H, m, CHH–O, THP

ring), 3.38 (1H, dt, CHH–O, chain), 2.03 (2H, m, C–CH₂–

CvC), 1.84 (1H, m), 1.70 (1H, m) 1.60–1.27 [21H, m,

(CH₂)₈þ(CH₂)₂þOH] ppm. ¹³C NMR: 133.4, 128.9 (alkene

C), 98.8 (acetal C), 67.7, 63.7, 62.3 (C–O), further

methylene ¹³C absorptions at 32.16, 30.77, 29.73, 29.51,

29.48, 29.42, 29.36, 29.11, 29.10, 26.20, 25.48, 19.65 ppm.

Mass spectrum (EI), peak match at M^þcalcd for C₁₈H₃₄O₃:

298.2508; found: 298.25085. Peak match at M285, calcd

for C₁₃H₂₅O₂: 213.1855; found: 213.1855.

1

35.17 g of product 7 (94%). H NMR (CDCl₃): d 4.50 (m,

1H, OCHO, 3.60 (m, 4H, 2£CH₂–O), 3.35 (t, 2H, CH₂Br),

1.40 (m, 22H) ppm. This material was used in the next step.

5.1.3. 13-(Tetrahydropyran-2-yloxy)-tridec-2-yn-1-ol 8.

Ammonia (250 mL) was condensed in a 500 mL 3-necked

ﬂask. The temperature was kept between 240 and 250 8C

with dry ice/i-PrOH and 0.025 g of Li metal was added to

the ammonia giving a blue color on dissolution. The

reaction mixture was kept under nitrogen and stirred

magnetically. After addition of a small amount of

Fe(NO₃)₃·9H₂O the formation of LiNH₂was catalyzed

and could be observed by the change of color (blue to grey).

Finely cut pieces of Li (2.17 g, 312 mmol) were added

slowly. The initially blue color on addition of the Li,

eventually turned into a white-greyish one. A solution of

propargyl alcohol (8.74 g, 156 mmol) in THF (40 mL) was

then added dropwise. The resulting mixture was stirred for

another hour after which a solution of the bromide 7

(16.71 g, 52 mmol) in THF (70 mL) was added. Stirring was

continued for 1.5 h at 240 8C and the reaction mixture was

left overnight at ambient temperature resulting in complete

evaporation of the ammonia. The residue was cautiously

treated with water (120 mL) and then extracted with ether

(3£80 mL). The organic layer was washed with water and

dried (MgSO₄). After work-up 15.3 g of crude material

(99.3%, purity according to GLC 92.2%) was obtained. This

crude material was pure enough for further use. A small

amount (5.0 g) was puriﬁed further via chromatography

over silica gel (eluent hexane–CH₂Cl₂1:1, followed by

CH₂Cl₂–AcOEt 1:1), giving 4.46 g of pure product

according to GC. IR (neat): n_max3390 (OH), 2280, 2220

(CuC) cm²¹. ¹H NMR (400 MHz): d 4.58 (1H, m, O–CH–

O), 4.24 (2H, s, CH₂OH), 3.87 (1H, m, CHH–O, THP-ether

ring), 3.72 (1H, m, O–CHH, alkyl chain), 3.51 (1H, m,

CHH–O, THP-ether ring), 3.39 (1H, m, O–CHH, alkyl

chain), 2.18 (2H, m, CH₂–Cu), 1.90 (1H, br s, OH), 1.85

(1H, m), 1.70 (1H, m), 1.61–1.48 (8H, m), 1.36–1.28 (12H,

m) ppm. ¹³C NMR: signals at 98.78, 86.41, 78.37, 67.64,

62.25, 51.26, 30.73, 29.68, 29.43, 29.37, 29.35, 29.02,

28.77, 28.56, 26.16, 25.47, 19.61 and 18.68 ppm. Mass

spectrum (EI) peak match for M^þ(abundancy.8%) calcd for

C₁₈H₃₂O₃, 296.2352; found: 296.2344. Peak match for

M^þ2CH₂OH (abundancy 3.5%) calcd for C₁₇H₂₉O₂

265.2168; found: 265.2166.

The by-product (the allene, 2-trideca-11,12-dienyloxy-

tetrahydro-pyran) was isolated by chromatography over

silica gel (CH₂Cl₂). IR (neat): n_max3040 (vCH₂), 1950

.

(CvCvC) cm²^{1 1}H NMR (90 MHz): d 5.08 (1H,

appearing as quintet, CH–CvC), 4.73–4.47 (3H, qþm,

O–CH–OþCvCvCH₂), 4.03–3.2 (4H, 2£CH₂, CH₂–O–

C), 2.16–1.07 [24H, m, CH₂–C¼þ(CH₂)₈þ(CH₂)₃] ppm.

5.1.5. (2R,3R)-2,3-Epoxy-13-(2-tetrahydropyran-yloxy)-

tridecan-1-ol 10. A solution of Ti(OiPr)₄(29.00 g,

102 mmol) in dichloromethane (650 mL) was cooled to

230 to 240 8C. ^D-(2)-Diethyl tartrate (21.03 g, 102 mmol)

dissolved in dichloromethane (100 mL) was added in one

portion. The resulting solution was stirred for 5 min at the

low temperature and then allylic alcohol 9 (29.80 g,

100 mmol) was added. After stirring for another 5 min a

solution of tert.-butyl hydroperoxide (50 mL, 4.1 mol) in

dichloroethane was added. The reaction mixture was left

overnight at 225 8C. A 10% aqueous tartaric acid solution

(300 mL) was added slowly and the mixture was stirred at

210 8C for 30 min and for 1 h at room temperature. A clear

yellow mixture was obtained, the organic layer was washed

with water (3£) and dried over MgSO₄. The solvent was

evaporated and the residue was dissolved in diethyl ether

(400 mL). The solution was cooled to 0 8C and stirred with

1 N NaOH (400 mL) for 30 min. The ether layer was

washed once with water and dried over Na₂SO₄. Evapor-

ation of the solvent and removal of remaining t-BuOOH in

vacuo gave the crude product (28.30 g). This material was

stored and puriﬁed in portions when needed.

Crude epoxy alcohol 10 (6.30 g) was chromatographed over

130 g of silica gel. Elution with EtOAc–hexane (2:1) gave

0.48 g of allene and 0.22 g of starting material. Elution with

EtOAc–hexane (1:1) gave 4. 33 g of pure epoxy alcohol as

an oil (68.7%). [a]_D²⁰¼þ18 (CHCl₃, c¼0.5). IR: n_max3420

5.1.4. 13-(Tetrahydropyran-2-yloxy)-tridec-2-en-1-ol 9.

A mixture of diethyl ether–THF (650 mL, ratio 7:8), was

added to LAH (7.6 g, 0.2 mol). The suspension was stirred

magnetically, kept under nitrogen and cooled below

265 8C. A solution of the alkynol 8 in diethyl ether–THF

(400 mL, 7:8) was added slowly; the temperature was kept

below 2 60 8C. After the addition the temperature of the

reaction mixture was allowed to rise to room temperature,

then the temperature was raised to 35–40 8C and maintained

at this level during 12 h. The reaction mixture was

decomposed by careful addition of water at 210 8C. The

colorless Al(OH)₃was ﬁltered off and the ﬁltrate was dried

(MgSO₄). Work-up gave the product as a slightly colored

viscous oil (29.86 g, quant.). This crude product could be

1

(OH) cm²¹. H NMR (400 MHz) (CDCl₃): d 4.57 (1H, m,

O–CH–O), 3.92–3.84 (2H, m, CHH–OHþCHH–O THP-

ring), 3.73 (1H, dt, O–CHH–CH₂, alkyl chain), 3.62 (1H,

dd, CHHOH), 3.38 (1H, m, CHH–O THP-ring), 3.38 (1H,

dt, O–CHH–CH₂, alkyl chain), 2.96–2.90 (2H, m, HC–

O–CH), 1.71 (1H, m), 1.63–1.28 (25H, m) ppm. ¹³C NMR:

d 98.87 (C acetal), 67.69 (O–CH₂alkyl chain), 62.34

(CH₂–O, THP-ring), 61.77 (CH₂–OH), 58.34 (CH₂CH,

epoxy), 55.99 (CH–O–CH–CH₂OH), further methylene C

absorptions at 31.55, 30.82, 29.76, 29.51, 29.47, 29.44,

29.36, 26.23, 25.92, 25.54, 19.71 ppm, one ¹³C absorption

5246

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

was missing due to overlap of two signals. MS, peak match

M^þ, calcd 314.2457; found: 314.2453.

m) ppm. ¹³C NMR: 190.45 CvO, 98.70 O–CH–O, 67.56

CH₂–O, 61.69 CH₂–O, 58.93 C–CvO, epoxide, 58.66

C–CH₂, epoxide, 50.63 CH–N₂, further methylene carbon

absorptions at 31.77, 31.15, 30.31, 29.94, 29.86, 29.80,

29.73, 29.63, 29.48, 26.79, 26.00, 19.75 ppm. MS, EI^þ352

(M^þ) 0.3%, 351 (M^þ21) 0.7%, 269 (M^þ283) 8.6%, 101

(C₅H₉O^þ₂) 71%, 85 (C₅H₉O^þ) 100%.

Another puriﬁcation of crude epoxy alcohol (13.25 g) gave

a product (9.90 g) with a purity of 95.6% according to GC

(75.2% yield calcd on allylic alcohol 9).

5.1.6. (2S,3R)-2,3-Epoxy-13-(2-tetrahydro-pyran-yloxy)-

tridecan-1-al. To a magnetically stirred solution of oxalyl

chloride (3.30 g, 26 mmol) in dichloromethane (50 mL),

cooled at 260 8C and kept under nitrogen, was added

dropwise a solution of DMSO (2.73 g, 35.0 mmol) in

dichloromethane (20 mL). A solution of the epoxy alcohol

10 (5.41 g, 17.2 mmol) in dichloromethane (30 mL) was

added and the mixture was stirred for 15 min. Then,

triethylamine (12 mL) was added and the mixture was

allowed to warm-up to room temperature. Water (100 mL)

was added and the two layers were separated. The

dichloromethane layer was washed once more with water,

and dried (MgSO₄). Evaporation of the solvent gave crude

epoxy aldehyde (5.4 g) as a yellow colored oil. Purity

according to GLC: 88.6%. NMR (90 MHz) (CDCl₃₎: d 9.0

(1H, d, CHvO), 4.55 (1H, m, O–CH–O), 4.0–3.0 (6H, m,

2£CH₂OþCH–O–CH), 2.0–1.0 [24H, m, (CH₂)₉þ(CH₂)₃]

ppm. This aldehyde was used immediately for the next step.

5.1.8. (4R)-4,14-Dihydroxy-tetradec-2-enoic acid ethyl

ester 13. A solution of epoxy diazomethyl ketone 12

(3.84 g, 10.9 mmol) in EtOH (1050 mL) was irradiated in a

Pyrex tube at 300 nm. The progress of the reaction was

followed by IR. After 4 h the diazo absorption had vanished,

and the solvent was evaporated leaving 3.12 g of an oil.

According to GLC this product was 80% pure. According to

NMR the product had lost its THP protecting group. IR

(CCl₄): n_max3620 (sharp, OH), 3500 (br, OH), 2940, 2860

(aliphatic CH₂), 1740 (EtO–CvO), 1660 (CvC), cm²¹. ¹H

NMR (90 MHz) (CDCl₃): d 6.95 (1H, dd, J¼16 Hz,

J¼5.5 Hz, HCvC–CvO), 6.02 (1H, dd, J¼16, 2 Hz,

CvCH–CvO), 4.17 (3H, q, J¼7 Hz, O–CH₂CH₃and

CH–OH), 3.60 (2H, t, J¼6 Hz, CH₂–OH), 2.92 (2H, m,

OH), 1.8–1.1 [21H, m, (CH₂)₉þCH₃] ppm. The product

was used in the next step right away.

5.1.9. (4R)-4,14-Dihydroxy-tetradecanoic acid ethyl ester

14. Ni(OAc)₂·4H₂O (1.43 g, 5.8 mmol), was suspended in

EtOH (50 mL) and transferred to a hydrogenation ﬂask. A

1 mol solution of NaBH₄(6 mL) in EtOH was added, giving

a black suspension. A solution of the g-hydroxy-a,b-

unsaturated ester 13 (3.12 g, when pure 10.9 mmol) in EtOH

(50 mL) was added. The device was connected to a

hydrogen supply and the uptake of hydrogen at normal

pressure was measured. After uptake of 225 mL of

hydrogen the reaction mixture was ﬁltered through celite.

The solvent was evaporated and the residue puriﬁed by

chromatography over silica gel (100 g). Elution with

hexane–EtOAc 3:1, which was gradually changed to 1:1

gave the desired compound 14 as colorless ﬂuffy needles

(1.77 g, 57%, calcd on epoxy diazomethyl ketone). After

recrystallization from hexane–CH₂Cl₂: mp 48–50 8C. IR

(CHCl₃): n_max3600, 3450 (OH), 1765 (sh, CvO), 1720

(CvO) cm²¹. ¹H NMR (CDCl₃₎: d 4.13 (2H, q, J¼7.5 Hz,

O–CH₂CH₃), 3.60 (3H, tþm, t J¼5 Hz, HO–CH₂,þCH–

OH), 2.43 (2H, t, J¼7.5 Hz, CH₂CvO), 2.2 (2H, b s,

2£OH), 1.8–1.1 [23H, m, (CH₂)₉þCH₂þCH₃) ppm.

Elemental analysis for C₁₆H₃₂O₄(288.431), calcd, C:

66.63; H 11.18%; found: C: 66.59; H: 11.05%.

5.1.7. (3S,4R)-1-Diazo-3,4-epoxy-14-(2-tetrahydropyran-

yloxy)-tetradecan-2-one 12. The crude epoxy aldehyde

(5.4 g, assumed 17.2 mmol) was dissolved in t-BuOH

(200 mL). Isobutene (75 mL) was added as chlorine

scavenger. At room temperature a solution of NaClO₂

(12 g, 133 mmol) and NaH₂PO₄(12 g, 100 mmol) in water

(120 mL) was added dropwise. The resulting mixture was

left overnight at room temperature. Evaporation of the more

volatile compounds gave an oily compound in the water

layer. The oil was extracted with diethyl ether (4£40 mL)

and the solution was dried over MgSO₄. Evaporation of the

solvent gave the glycidic acid 11 as a crude oil in

quantitative yield. This product was used for the next

reaction immediately.

To an ice-cooled solution of the crude glycidic acid (5.37 g)

in diethyl ether (60 mL) was added iso-butyl chloroformate

(2.60 g, 19.0 mmol). This was followed by addition of a

solution of Et₃N (2.26 g, 22.4 mmol) in diethyl ether

(20 mL). After stirring for 30 min, the precipitated salt

was ﬁltered off and the ﬁltrate was added to an ethereal

diazomethane solution (143 mL, 0.3 mol) that was cooled at

230 8C. The mixture was left overnight at room tempera-

ture and excess diazo methane was removed by a stream of

nitrogen. Some additional Et₃N·HCl salt was removed by

ﬁltration and the ﬁltrate was concentrated. A yellow oil

remained (5.88 g), that was puriﬁed by chromatography

over 90 g of silica gel. Elution with hexane–EtOAc 4:1

yielded the pure epoxy diazomethyl ketone 12 (4.00 g, 68%

calculated on the epoxy alcohol) as a yellow oil.

[a]²_D⁰¼243.1 (CHCl₃, c¼1). IR (neat): n_max3115 (CHv),

2940, 2860 (aliph CH₂), 2110 (vNvN), 1640 (CvO)

cm²¹. ¹H NMR (C₆D₆, 400 MHz): d 4.75 (1H, s, CHvN₂),

4.62 (1H, m, appears as t with J¼3.0 Hz, O–CH–O), 3.87–

3.82 (2H, m, CH₂–O), 3.44–3.35 (2H, m, CH₂O), 3.04 (1H,

d, J¼2 Hz, O–CH–CvO), 2.56 (1H, bs, CH₂–CH–O,

epoxide), 1.79 (1H, m), 1.65–1.61 (4H, m) 1.42–1.02 (19H,

5.1.10. (5R)-5-(10-Hydroxydecyl)-dihydrofuran-2-one

15. A solution of the dihydroxy ester 14 (1.40 g,

4.85 mmol) in benzene (80 mL) containing a small amount

of TsOH, was heated at reﬂux for 2 h. The reaction was

monitored by TLC. The reaction mixture was washed with

aq. NaHCO₃and water. Work-up gave compound 15

(1.18 g, quant.) as colorless crystals (mp 80–81 8C).

[a]²_D⁰¼þ23.9 (CHCl₃, c¼0.8). IR (CHCl₃): n_max3600,

1

3470 (OH), 1765 (CvO), cm²¹. H NMR (400 MHz) d

4.48 (1H, m, CHOCvO), 3.64 (2H, t, J¼6.5 Hz, CH₂OH),

2.52 (2H, m, CH₂–CvO), 2.32 (1H, m, CHH–CH₂–

CvO), 1.71 (1H, m, CHH–CH₂–CvO)1.60–1.23 [19H,

m, (CH₂)₉þOH] ppm. ¹³C 177.15 CvO, 81.00 CH–O,

63.01 CH₂OH, CH₂peaks at 35.56, 32.76, 29.52, 29.47,

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5247

29.34, 29.26, 27.96, 26.18, 25.50, 25.42 and 25.16 ppm.

Elemental analysis for C₁₄H₂₆O₃(242.360): calcd, C: 69.38;

H: 10.81%; found: C: 68.66; H: 10.81%.

5.1.14. (3R/S,5R)-5-Dec-9-enyl-3-{[(4S)-2,2-dimethyl-

[1,3]dioxolan-4-yl]-hydroxy-methyl}-dihydrofuran-2-

one 19/20. To a LDA solution [prepared from iPr₂NH

(303 mg, 3 mmol) and BuLi (2 mL of 1.6 molar sol. in

hexane)] in THF (5 mL) while kept at 270 8C, was slowly

added a solution of the lactone 15 (448 mg, 2 mmol) and

DMPU (320 mg, 2 mmol) in THF (5 mL).The reaction

mixture was stirred for 15 min at this temperature and then a

solution of (4R)-2,2-dimethyl-[1,3]-dioxolane-4-carbalde-

hyde 18 (781 mg, 6.0 mmol) in THF (5 mL) was added in

one portion. The reaction mixture was kept at 270 8C for

6 h. The reaction was quenched with satd aqueous NH₄Cl

and the mixture was extracted with ether. After drying

(MgSO₄), work-up gave an oil that was puriﬁed by column

chromatography over 40 g of silica gel. Elution with

CH₂Cl₂gave starting material (40 mg, 9%). With diethyl

ether the product (650 mg, 91%) was obtained as a mixture

of two isomers 19/20 (according to GLC). By careful

crystallization (diethyl ether–hexane) one of the alcohols

was isolated in pure form (mp 60–62 8C). A mass spectrum

(CI) showed the M^þ1peak at 355. IR (CCl₄): n_max3600

(OH, w), 3500–3400 (OH, m, b), 3080 (vCH, w), 2990,

2930, 2860 (aliphatic hydrogens) 1760 (CvO, s), 1640

(CvC, w), 1380, 1370 (typical absorptions of the dioxolane

function, cm²¹. ¹H NMR (90 MHz FT) (CDCl₃): d 6.2–5.6

(1H, m, HCvCH₂), 5.15–4.8 (2H, m,vCH₂), 4.75–4.4

(1H, m, CHOCvO), the remaining part of the spectrum

contained several signals, however, the resolution was to

poor to allow a reliable assignment. The mixture of 19/20

was used as such in the next step.

5.1.11. (5R)-5-[10-(2-Tetrahydropyran-yloxy)]-decyl-

dihydrofuran-2-one 37. The alcohol function of compound

15 was reprotected as follows: to a solution of alcohol 15

(435 mg, 1.8 mmol) in CH₂Cl₂(10 mL) cooled at 0 8C, a

solution of dihydropyrane (160 mg, 1.9 mmol) in CH₂Cl₂

(2 mL) was added. A small amount of TsOH was added as

catalyst. After 1 h at room temperature the solution was

washed with satd bicarbonate solution and dried over

MgSO₄. Work-up and chromatography over silica gel

(hexane–EtOAc 2:1) gave product 37 as a slightly yellow

oil (500 mg, 85%). [a]²_D⁰¼þ20.6 (CHCl₃, c¼0.6). ¹H NMR

(400 MHz) (CDCl₃): d 4.57 (1H, m, O–CH–O), 4.48 (1H,

m, CH–O–CvO), 3.87 (1H, m, CHH–O THP-ring), 3.73

(1H, m, CHH–O, chain), 3.52 (1H, m, CHH THP-ring),

3.38 (1H, m, CHH–O, chain), 2.52 (2H, m, CH₂–CvO),

2.32 (1H, m, CHH–CH₂–CvO), 1.84 (1H, m, CHH–

CH₂–CvO), 1.8–1.28 [24H, m, (CH₂)₃þ(CH₂)₉)] ppm.

¹³C NMR: 177.05 CvO, 98.84 O–CH–O, 80.95 CH–O–

CvO, 67.64 O–CH₂chain, 62.31 O–CH₂THP-ring, other

CH₂peaks at 35.55, 32.76, 30.77, 29.71, 29.47, 29.39,

29.37, 29.27, 28.78, 27.95, 26.18, 25.49, 25.16, 19.67 ppm.

5.1.12. (5R)-5-[10-(2-Nitro-phenylselanyl)-decyl]-

dihydrofuran-2-one 16. To a magnetically stirred solution

of alcohol 15 (754 mg, 3.11 mmol) and o-nitrophenyl-

selanyl cyanide (1181 mg, 5.20 mmol) in THF (25 mL) was

added at room temperature a solution of Bu₃P (1296 mL,

5.20 mmol) in THF (10 mL). The reaction was followed by

TLC. After 2 h the reaction was complete and the deeply

reddish-brown colored solution was concentrated. The

residue was subjected to column chromatography over

silica gel, eluent EtOAc–hexane 1:3!1:2. This gave yellow

5.1.15. Reaction of alcohol 19/20 with N,N⁰-thiocar-

bonyldiimidazole. At room temperature. To a solution of

the alcohol mixture of ₀19/20 (350 mg, 1.0 mmol) in THF

(5 mL) was added N,N -thiocarbonyldiimidazole (350 mg,

2 mmol) at room temperature. The reaction mixture was

stirred for 6 h at this temperature. The solvent was

evaporated under reduced pressure and the residue was

chromatographed over 30 g of silica gel with hexane–

EtOAc 2:1. A small amount of conjugated unsaturated

product 23þ24 (20 mg, 6%) was eluted ﬁrst, followed by

fractions with starting material (73 mg, 21%), and the

desired thiocarbonyl-imidazolide mixture 21 (an oil,

298 mg, 64. IR (CCl₄): n_max3160 (w), 3145 (w),

3075(w,vCH), 2995, 2925, 2860 (aliphatic CH₂), 1770

(CvO), 1640 (vCH₂), 1380, 1370 (typical absorptions of

crystals (1186 mg, 93%) of compound 16. IR (CHCl₃):

1

n

_max1760 (CvO), 1590 (Ar), 1500, 1335 (NO₂). H NMR

(CDCl₃): d 8.23 (1H, d, J¼8.0 Hz, CHvC–NO₂), 7.6–7.0

(3H, m, ArH), 4.48 (1H, m, CH–OCvO), 2.90 (2H, t,

J¼7.5 Hz, CH₂Se), 2.7–1.1 [22H, m, (CH₂)₉þ(CH₂)₂] ppm.

This material was processed immediately in the next step.

5.1.13. (5R)-5-(Dec-9-enyl)-dihydrofuran-2-one 17. To a

solution of the selenide 16 (532 mg, 1.25 mmol) in THF

(15 mL) at a temperature of 5 8C was added hydrogen

peroxide (1.5 mL 30%, 13 mmol). The reaction mixture was

stirred for 3.5 h at room temperature. At this point very little

starting material remained. After 5 h the starting material

was consumed completely. Then, hexane (50 mL) was

added. A satd aqueous solution of NaHCO₃(50 mL) was

added and the mixture was stirred vigorously. The aqueous

layer was extracted twice with hexane (40 mL). The

combined organic layers were dried over MgSO₄. Work-

up and puriﬁcation by chromatography over silica gel

(30 g), eluent EtOAc–hexane 1:5, gave compound 15

(190 mg, 68%) as a yellow oil. The yellow color is due to

a small trace of a selenide contaminant. According to GLC

the purity was .95%. IR (neat): n_max3060 (CvCH), 1770

(CvO), 1640 (CvC) cm²¹. ¹H NMR: d 6.23–5.33 (1H, m,

H₂CvCH), 5.3–4.7 (2H, m, H₂CvC), 4.47 (1H, m, CH–

OCvO), 2.7–1.1 [20H, m, (CH₂)₈þ(CH₂)₂] ppm. This

material was coupled with 18 without further puriﬁcation.

1

dioxolane), cm²¹. H NMR (CCl₄): d 8.2, 7.6, 6.9, 3£s of

imidazole ring. The proton at the carbon with the

imidazolide substituent was present as a multiplet at

5.8 ppm. No further assignments were made.

In THF at reﬂux temperature. A solution of the alcohol

mixture 19/20 (177 mg, 0.5 mmol) and N,N⁰-thiocarbonyl-

diimidazole (178 mg, 1 mmol) in dry THF (10 mL) was

reﬂuxed for 3 h. Still some starting material 19/20 was

left. Another portion of N,N⁰-thiocarbonyldiimidazole

(45 mg, 0.25 mmol) was added and heating at reﬂux

temperature was continued for 1 h. The solvent was

evaporated and the residue was chromatographed over

silica gel (hexane–EtOAc 2:1). The conjugated unsaturated

products 23þ24 were obtained ﬁrst (37 mg, 22%),

followed by starting material (34 mg, 20%) and the

desired product 21 as an oil (128 mg, 55%). The

5248

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

imidazolides 21 were used in the next step without further

characterization.

Z-isomers 24 and 23, in equal amounts according to GC.

Another chromatographic puriﬁcation with hexane–EtOAc

as eluent gave also a small amount of E-isomer 24. IR

(CCl₄): n_max3070 (w, vCH), 2990, 2920, 2850 (s, aliphatic

CH₂and CH₃), 1750 (s, CvO), 1670 (w, CvC–CvO),

5.1.16. (3R/S,5R)-5-Dec-9-enyl-3-[(4S)2,2-dimethyl-[1,3]-

dioxolan-4-yl]-methyl-dihydrofuran-2-one 22. To

a

solution of thiocarbonyldiimidazoles 21 (90 mg,

0.19 mmol) in toluene (10 mL) was added freshly prepared

tributyltin hydride (125 mg, 0.43 mmol). The solution was

slowly heated to reﬂux temperature, at which point the

reaction was complete. After removal of the solvent, the

residue was puriﬁed by ﬂash chromatography (EtOAc–

hexane 1:4) over silica gel. This afforded product 22 as an

oil (45 mg, 70%). According to GLC two isomers were

present. IR (neat): n_max3070, 2980, 2920, 2850, 1770, 1640,

1640 (w, CvC), 1380, 1370 (dioxolane absorptions) cm²¹

.

¹H NMR (CDCl₃): d 6.66 (1H, dt, J¼8 Hz, J¼1.8 Hz,

OvC–CvCH), 6.1–5.6 (1H, m, H₂CvCH), 5.15–4.8

(2H, m,vCH₂), 4.9–4.3 (2H, HC–O–CvO,vC–CH–O),

4.3–4.1 (1H, s, HHC–O), 3.9–3.6 (1H, m, HHC–O), 3.25–

2.45 (2H, dq, CH₂lactoneþallyl coupling), 2.2–1.9 (2H, m,

vCH–CH₂), 1.95–1.0 (the remaining protons as broad

multiplet) ppm.

1380, 1370 cm²¹. H NMR (90 MHz) (CDCl₃): d 6.1–6.5

5.1.19. (3S,5R,4⁰S)-5-Decyl-3-(2⁰,2⁰-dimethyl-[1⁰,3⁰]-

dioxolan-4⁰-yl)-methyl-dihydrofuran-2-one 27. A mixture

of the two isomeric alkenes 23þ24 (20 mg) was dissolved in

EtOH (10 mL) and hydrogenated over Pd/C catalyst at room

temperature at normal pressure. After 2 h the mixture was

ﬁltered over hyﬂo and the ﬁltrate concentrated. This gave a

single compound 27 (18 mg), according to GC, as an oil that

crystallized on standing. Recrystallization from hexane

gave a solid, mp 38–41 8C. (mp of acetonide of rubrenolide

47–48 8C). IR (CCl₄): n_max2995, 2940, 2860 (s, CH₂, CH₃),

1770 (s, CvO), 1470, 1455 (m), 1380, 1370 (s), 1170 (s)

cm²¹. ¹H NMR (90 MHz, FT) (CDCl₃): d 4.5–4.2 (1H, m,

CHOCvO), 4.2–4.0 (2H, m, CH₂O), 3.7–3.4 (1H, m, CH–

O–CH₂–O), 2.9–2.35 (2H, m), 2.3–2.0 (1H, m), 1.9–1.2

(remaining protons as multiplet, with singlets of CH₃at 1.38

and 1.32), 0.88 (3H, broad triplet, CH₂–CH₃) ppm.

1

(1H, m, HCvCH₂), 5.2–4.85 (2H, m, CH₂¼CH), 4.80–4.0

(2H, m, 2£CH–O), 4.0–3.3 (2H, m, CH₂–O) ppm, the

remaining signals could not be assigned. This material was

used as such in the next deprotection step.

5.1.17. (3S,5R,2⁰S)-5-Dec-9⁰⁰-enyl-3-(2⁰,3⁰-dihydroxy-pro-

pyl)-dihydrofuran-2-one 25 and (3R,5R,2⁰S)-5-dec-9⁰⁰-

enyl-3-(2⁰,3⁰-dihydroxy-propyl)-dihydrofuran-2-one 26.

Product 22 (45 mg) was dissolved in a small amount of

MeOH (5 mL) and TosOH (50 mg) was added. After

stirring for 1 h at room temperature hydrolysis was

complete. The solvent was evaporated and the residue

dissolved in diethyl ether. After washing with bicarbonate

solution, the organic phase was dried (MgSO₄), ﬁltered and

the solvent evaporated. The residue was puriﬁed by ﬂash

chromatography over silica gel (EtOAc–hexane 2:1). This

gave 20 mg of product (61%). It was pure according to TLC

and GLC. However, a long melting range was observed

5.1.20. Epi-dihydro-rubrenolide 28. Acetonide 27 (5 mg)

was treated with p-TsOH in MeOH (1 mL). Almost

immediate hydrolysis of the acetonide function took place.

A small amount of satd aq. NaHCO₃soln was added and the

mixture was extracted with ether. After drying and removal

of the solvent a solid was obtained. Recrystallization from

ether–hexane gave ﬁne needles, mp 94–96 8C. IR (KBr):

n_max3450 (OH), 2995, 2945, 2860 (CH₂, CH₃), 1745

1

starting at 65 8C. A 400 MHz H NMR spectrum in CDCl₃

showed signals of two products. Compound 25: d 5.80

(m), 4.94 (m), 4.38 (m), 3.98 (m), 3.68 (m), 3.49(m), 2.84

(m), 2.54 (m), 2.03–1.28 (m) ppm. Compound 26: d 5.80

(m), 4.94 (m), 4.60 (m), 3.82 (m), 3.65 (m), 3.55 (m), 2.91

(m), 2.18 (m), 2.03–1.28 (m) ppm. The ratio of 25/26 was

2:3.

1

(CvO) cm²¹. H NMR (400 MHz) (CD₃OD): d 4.46 (1H,

m, CHO–CvO), 3.88 (1H, m, HO–CH–CH₂–OH), 3.51

(2H, m, CH₂–OH), 2.90 (1H, m, CH–CvO), 2.60 (1H, m,

lactone ring HHC–C–O–CvO), 2.06 (1H, m, CHH of

dihydroxy-propyl side chain), 1.74 (1H, m, CHH of

dihydroxy-propyl side chain), 1.7–1.3 (19H, m), 0.9 (3H,

t, J¼7 Hz, CH₂–CH₃) ppm.

5.1.18. Reaction of thiocarbonylimidazolide 21 with

tributyltin hydride in toluene at reﬂux temperature. A

solution of the imidazolide 21 (crude 840 mg, 1.8 mmol) in

toluene (20 mL) was added dropwise to a solution of tributyl

tinhydride (1.16 mL, 4.0 mmol) in toluene (20 mL) which

was heated at reﬂux. The mixture was heated at reﬂux for

another hour. After cooling, the solvent was evaporated

under reduced pressure and the residue was chromato-

graphed over silica gel (45 g). Eluting with hexane–EtOAc

4:1, followed by hexane–EtOAc 2:1. The ﬁrst fraction

(50 mg, 8%) was the pure conjugated Z-alkene 23. IR

(CCl₄): n_max3070 (w, vCH), 2990, 2920, 2850 (s, aliphatic

CH₂and CH₃), 1755 (s, CvO), 1675 (w, CvC conjugated),

1

For comparison the H NMR spectrum of dihydrorubreno-

lide obtained through hydrogenation of natural rubrynolide

1

is given. H NMR (400 MHz) (CD₃OD): d 4.40 (1H, m,

CH–O–CvO), 3.60 (1H, m, HO–CH–CH₂–OH), 3.49

(2H, dd, CH₂–OH), 2.94 (1H, m, CH–CvO), 2.54 (1H,

ddd, lactone ring HHC–C–O–CvO, a H), 1.92 (1H, ddd,

CHH of dihydroxy-propyl side chain), 1.70 (1H, m, CHH of

dihydroxy-propyl side chain), 1.60 (1H, m, lactone ring

HHC–C–O–CvO, bH), 1.6–1.25 (remaining protons),

0.90 (3H, t, J¼7 Hz) ppm. This spectrum clearly differs

from that of 28.

1640 (w,vCH₂) 1380, 1370 (dioxolane absorptions) cm²¹

.

¹H NMR: d 6.20 (1H, dt, J¼8 Hz, J¼1.5 Hz, OvC–

CvCH), 6.05–5.50 (2H, m, CH₂vCH, CvCH–CH–O),

5.1–4.75 (2H, m,vCH₂), 4.6–4.0 (2H, m, CHO–CvO,

CHH–O), 3.75–3.40 (1H, m, CHH–O), 3.15–2.15 (2H, dq,

J¼18 Hz, J¼7.0 Hz, and an allylic coupling of 1.5 Hz,

lactone CH₂), 2.15–1.8 (2H, m,vCH–CH₂), 1.8–0.6 (rest

of protons, m) ppm. The next fraction gave a colorless oil

(339 mg, 56%). which consisted of a mixture of the E and

5.1.21. Tridec-2-yn-1-ol.³¹1-Bromotridecane (20.0 g,

91 mmol), propargylic alcohol (14.84 g, 264 mmol) and

LiNH₂(from 3.6 g of Li, 514 mgat) in ammonia (100 mL)

was brought in reaction as described for compound 8. After

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5249

1.5 h DMSO (15 mL) was added and the reaction mixture

was left overnight After work-up the oil obtained was

puriﬁed by destillation b.p. 94–95 8C (0.05 Torr), yield

16.0 g (90%). The oil solidiﬁed on standing, lit. mp 37–

38 8C. IR (CCl₄): n_max3620 (OH), 2280, 2220 (alkyn)

added isobutyl chloroformate (5.66 g, 41.5 mmol), followed

by triethylamine (6.26 g, 61.8 mmol). After stirring for 1 h

at the indicated temperature the precipitate was ﬁltered off

under a stream of nitrogen. The ﬁltrate was added to a

solution of diazomethane in diethyl ether (300 mL,

concentration of CH₂N₂: 0.3 mmol/mL). After one night

at room temperature the reaction mixture was ﬂushed with

nitrogen to remove excess diazomethane. Residual Et₃-

N·HCl was ﬁltered off and the solvent was evaporated

leaving a black residue. Flash chromatography (EtOAc–

hexane 8:1) over silica gel gave the epoxy diazomethyl

ketone as a yellow solid (4.93 g, yield calcd on epoxy

alcohol 47%), mp 41–42 8C (hexane–diethyl ether).

Elemental analysis: calcd for C₁₄H₂₄N₂O₂(252.357): C

66.63, H 9.59, N 11.10%; found: C 66.38, H 9.60, N

10.83%_;[a]²_D⁰¼265.7 (CHCl₃, c¼1). IR (CCl₄): n_max3120

1

cm²¹. H NMR (CDCl₃): d 4.2–4.4 (2H, bs, –CH₂OH),

2.1–2.3 (2H, m, CH₂Cu), 1.2–1.7 [16H, m, (CH₂)₈)], 0.9

(3H, t, CH₂CH₃) ppm.

5.1.22. E-Tridec-2-ene-1-ol. To a suspension of LiAlH₄

(10.0 g, 262 mmol) in diethyl ether (75 mL) that was cooled

with an ice-salt bath, a solution of tridec-2-yn-1-ol (25.7 g,

130 mmol) in diethyl ether (75 mL) was added over a period

of 2.5 h. Then THF (170 mL) was added and the mixture

was heated at reﬂux for 2.5 h. After cooling in ice a few mL

of water were added carefully to hydrolyse the Li/Al salts. A

white solid was formed and the mixture was dried over

MgSO₄. Work-up gave a light-yellow oil that solidiﬁed on

standing at 0 8C (23.45 g). This alkene was contaminated

with the allene (15%) according to GC and it was used as

such for the next reaction.

1

(HCv), 2115 (CvNvN), 1645 (CvO) cm²¹. H NMR

(300 MHz) (CDCl₃): d 5.41 (1H, s, HCvN₂), 3.15 (1H, d,

J¼1.4 Hz, O–CH–CvO), 2.90 (1H, m, –CH₂CH–O),

1.6–1.2 [(18H, m, (CH₂)₉], 0.81 (3H, t, J¼6.9 Hz, CH₃)

ppm. ¹³C NMR: 147.80 CvO, 59.26 CH₂C–O, 58.75

O–CH–CvO, 51.96 CHvN₂, CH₂signals at 31.85, 31.66,

29.53, 29.45, 29.41, 29.27, 29.21, 25.63, 22.64, CH₃at

14.08 ppm.

5.1.23. (2R,3R)-2,3-Epoxy-tridecane-1-ol 29.³²To

a

˚

magnetically stirred suspension of molsieves (1.51 g, 4 A)

in dichloromethane (175 mL), while kept at 220 8C, D(2)-

DET (625 mg, 3.0 mmol) and Ti(OiPr)₄(.72 g, 2.5 mmol)

were added. After slow addition (1 h) of TBHP (25 mL of a

4.08 molar solution in dichloroethane, 100 mmol) the

slightly yellow solution was stirred at 220 8C for 30 more

minutes. A solution of tridec-2-ene-1-ol (10 g, 50.5 mmol)

in dichloromethane was added over a period of 1 h. The

reaction mixture was kept at 220 8C for 3 h. The

temperature was allowed to rise to 0 8C and the reaction

mixture was then poured into a solution of ferrosulfate

(16 g, 60 mmol) and tartaric acid (5.0 g, 30 mmol) in water

(50 mL). After stirring vigorously for 15 min the two layers

were separated. The water layer was extracted with ether

(2£75 mL) and the organic layers were stirred with a

solution of NaOH (30%, 25 mL, satd with NaCl, kept at

0 8C). After stirring at 0 8C for 45 min, water (250 mL) was

added and the layers were separated. The water layer was

extracted once more with ether (150 mL) and the organic

phase was dried over MgSO₄. Work-up afforded a colorless

solid (8.93 g, 77%). Chromatography (EtOAc–hexane 1:3)

over silica gel gave the pure compound (7.5 g, 69%),

[a]_D²⁰¼þ25 (CHCl₃, c¼1). Lit.³²[a]_Dof the 2S,3S

compound 224 (CHCl₃). IR (CCl₄): n_max3600 (OH)

5.1.25. Ethyl (4R)-4-acetoxy-tetradec-2-enoate 31. A

solution of (3S,4R)-1-diazo-3,4-epoxy-tetradecan-2-one 30

(2.36 g, 9.35 mmol) in abs ethanol (800 mL) was irradiated

for 2.5 h. The reaction was monitored by taking IR spectra.

Evaporation of the solvent gave a yellow oil (2.50 g). A

mixture of thus obtained ethyl (4R)-4-hydroxy-tetradec-2-

enoate (5.09 g, 18.7 mmol), pyridine (6.9 g, 85 mmol),

acetic anhydride (6.0 g, 60 mmol) and a small amount of

DMAP was stirred at room temperature for 18 h. The

mixture was poured in ice-water (100 mL) and extracted

with ether (3£100 mL). After drying (MgSO₄), the solvent

was evaporated and the residue was puriﬁed by ﬂash

chromatography (hexane–EtOAc 20:1), giving 31 as a

colorless oil (2.89 g, 50%). IR (CCl₄): n_max1740 (CvO),

1720 (vC–CvO), 1660 (CvC) cm²¹. ¹H NMR (90 MHz)

(CCl₄): d 6.8 (1H, dd, J¼15, 4.5 Hz, EtO–C(vO)–

CHvCH–CH), 5.9 (1H, d, J¼15 Hz, OvC–CH), 5.5–

5.2 (1H, m, CH–O–C(CvO)–CH₃), 4.1 (2H, q, J¼6.7 Hz,

O–CH₂CH₃), 2.0 (3H, s, OvC–CH₃), 1.7–1.1 [(21H, m,

(CH₂)₉þCH₃CH₂–O)], 0.9 (3H, t, CH₂–CH₃) ppm.

A 400 MHz spectrum was run and the ee was determined

(91%) with the aid of the optical shift reagent Eu(hfc)₃. The

chemical shifts found of the 400 MHz spectrum matched

approximately with the 90 MHz spectrum; the values were

as follows: 6.84 (1H, d d, J¼15.7 Hz, J¼5.4 Hz), 5.93 (1H,

d, J¼15.7 Hz), 4.20 (1H, m), 4.20 (2H, q, J¼7.2 Hz), 2.09

(3H, s), 1.33–1.25 (21H, m), 0.88 (3H, t, J¼7.2 Hz) ppm.

cm²^{1 1}H NMR (90 MHz) (CCl₄): d 4.40–3.4 (2H, m,

.

CH₂–OH), 3.0–2.8 (2H, m, HC–CH), 1.8–1.2 [(18H, m,

(CH₂)₉], 0.9 (3H, t, CH₃) ppm.

5.1.24. (3S,4R)-1-Diazo-3,4-epoxy-tetradecan-2-one 30.

To a mixture of (2R,3R)-epoxy-tridecan-1-ol 29 (8.89 g,

41.5 mmol) in CCl₄(80 mL), acetonitrile (80 mL) and water

(120 mL) was added NaIO₄(29.7 g, 138.8 mmol) and

RuCl₃·3H₂O (240 mg, 2.2 mol%). The brown mixture was

stirred for 2.5 h at room temperature. According to TLC the

reaction was complete. Dichloromethane (350 mL) and

water (200 mL) were added and the layers were separated.

The water layer was extracted with dichloromethane

(3£100 mL). After drying (MgSO₄) and evaporation of

the solvent a black residue was residing. This was dissolved

in diethyl ether (500 mL). To the cooled (2 8C) solution was

5.1.26. (5R)-5-Decyl-tetrahydrofuran-2-one 32. To a

suspension of palladium dichloride (31 mg, 0.17 mmol) in

EtOH (15 mL), NaBH₄(15 mg, 0.36 mmol) was added. The

mixture was stirred for 15 min and ethyl (4R)-4-acetoxy-

tetradec-2-enoate 31 (1.08 g, 3.46 mmol) was added. In a

closed system the mixture was kept in a hydrogen

atmosphere applying a small over-pressure. The theoretical

amount of hydrogen was absorbed in 90 min. The mixture

was ﬁltered over hyﬂo and the ﬁltrate was evaporated. The

5250

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

residue was chromatographed over silica gel (EtOAc–

petroleum ether 1:20). Ethyl (4R)-4-acetoxy-tetradecanoate

was obtained as a colorless solid (827 mg, 76%). IR (CCl₄):

0 8C a few crystals of DMAP and methanesulfonyl chloride

(2.0 mL) were added. After 1 h at room temperature, the

dichloromethane was evaporated, the residue was dissolved

in ether (30 mL). The solution was washed with water and

an aqueous solution of CuSO₄. The ether solution was dried

over MgSO₄and the solvent was evaporated leaving a

yellow oil (825 mg). To a solution of this crude mesylate

(825 mg, 1.9 mmol) in dichloromethane (20 mL) DBU

(580 mg, 3.8 mmol) was added. The reaction was monitored

by TLC. After 2 h the reaction was completed. The

dichloromethane solution was washed with aq. bicarbonate

and dried (MgSO₄). After evaporation of the solvent a

mixture of cis and trans alkene 34 was obtained (620 mg,

96%). IR (CCl₄): n_max1760 (CvO), 1685 (CvC), 1380,

1370 (dioxolane absorptions) cm²¹. This material was then

subjected to hydrolysis.

n

_max1740, 1730 cm²¹(CvO). ¹H NMR (90 MHz): d 4.85

(1H, m, CH–O–CvO), 4.1 (2H, q, J¼6.7 Hz, CH₃CH₂–

O), 2.3 (2H, t, CH₂C(vO)–O), 2.0 (3H, s, O–C(vO)–

CH₃), 2.0–1.7 (2H, m, CH₂–CH₂–C(CvO)–OEt),

1.7–1.1 [21H, m, (CH₂)₉þO–CH₂–CH₃), 0.9 (3H, t,

J¼6.0 Hz, CH₃-alkyl) ppm.

To a solution of (4R)-4-acetoxy-tetradecanoate (980 mg,

3.12 mmol) in EtOH (20 mL) was added KOH (0.8 g,

14.2 mmol). The solution was heated at reﬂux for 15 min.

Then the solvent was evaporated and the residue was

dissolved in a small amount of water. After one extraction

with ether the aqueous layer was acidiﬁed and extracted

with ether (3£). These ether extracts were concentrated and

the residue was dissolved in benzene (80 mL). A small

amount of TsOH was added and the solution was heated at

reﬂux and water was removed azeotropically. The benzene

solution was washed with aq. NaHCO₃solution. After

drying, evaporation of the solvent gave the product as a pale

yellow solid (711 mg, quant.). Recrystallization from

heptane gave colorless crystals, mp 35–36 8C, [a]²_D⁰¼þ32

(CHCl₃, c¼1.1), lit.²¹mp 36–37 8C, [a]²_D⁶¼þ29.95

(CHCl₃, c¼2.7). Elemental analysis, calcd for C₁₄H₂₆O₂

(226.359): C 74.29, H 11.58%; found: C 74.33, H 11.39%.

5.1.29. (5R)-5-Decyl-3-[(2R)-2,3-dihydroxypropylidene]-

dihydro-furan-2-one 35. The alkene mixture 34 (560 mg,

1.66 mmol) was dissolved in MeOH (20 mL) and a small

amount of TsOH was added. After stirring for 3 h aq.

NaHCO₃solution was added and the methanol was

evaporated. The residue was treated with water and the

mixture was extracted with ether. Usual work-up gave 35 as

a colorless solid (380 mg, 77%) which was hydrogenated

right away.

IR (CCl₄): n_max1780 (CvO) cm²¹. H NMR (300 MHz)

5.1.30. (3S,5R,2⁰R)-5-Decyl-3-(2⁰,3⁰-dihydroxy-propyl)-

dihydrofuran-2-one, (dihydro rubrenolide) 36. The

lactone 35 (380 mg, 1.28 mmol) was dissolved in EtOH

(10 mL) and hydrogenated over Pd/C as catalyst at normal

pressure at room temperature. After shaking for 3 h in a

hydrogen atmosphere the uptake of hydrogen had

stopped. Filtration over hyﬂo and evaporation of the solvent

gave a pure residue 36 (320 mg, 84%), mp 105–107 8C (lit.³

106–107 8C), [a]_D²⁰¼þ21.8 (CHCl₃, c¼1) (lit.³þ22,

CHCl₃). IR and NMR (400 MHz) spectra of 36 were in

complete agreement with the spectra obtained from

dihydrorubrenolide that was prepared by hydrogenation of

authentic rubrynolide obtained from Professor Ollis.²²

1

(CDCl₃): d 4.5 (1H, m, CHOCvO), 2.56–2.50 (2H, m,

CH₂CvO), 2.37–2.26 (1H, m, CHCH–O–CvO), 1.9–

1.26 [19H, m, CHCH–O–CvO, (CH₂)₉)], 0.88 (3H, t,

J¼6.9 Hz, CH₃) ppm. ¹³C NMR: 177.25 CvO, 81.02 CH–

O, CH₂peaks at 35.54, 31.85, 29.53, 29.48, 29.42, 29.30,

29.27, 28.83, 27.97, 25.18, 22.64, CH₃at 14.07 ppm.

5.1.27. (3R/S,5R)-5-Decyl-3-{[(4R)-2,2-dimethyl-

[1,3]dioxolan-4-yl]-hydroxy-methyl}-dihydrofuran-2-

one 33. To a stirred solution of di-isopropyl amine (505 mg,

5.0 mmol) in THF (10 mL) kept at 0 8C, was added a

solution of BuLi in hexane (3.12 mL, 1.6 N). The mixture

was cooled at 278 8C and a solution of the lactone 32

(567 mg, 2.5 mmol) in THF (5 mL) was added dropwise,

followed by DMPU (390 mg, 2.2 mmol) in THF (5 mL).

The solution was stirred for 15 min and (4S)-2,2-dimethyl-

[1,3]-dioxolane-4-carbaldehyde 38 (1.0 g, 7.7 mmol) in

THF (5 mL) was added. The reaction mixture was kept at

the low temperature for 4 h and a satd aq. solution of

ammonium chloride (25 mL) was added. The organic layer

was collected and the water layer was extracted with ether

(3£25 mL). After drying over Na₂SO₄, the solvents were

evaporated to give a pale yellow oil (730 mg). Chromato-

graphy over silica gel (EtOAc–petroleum ether 1:5 gave the

product consisting of mainly two isomers (0.56 g, 63%).

Some starting material (84 mg, 15%) was also obtained. IR

(CCl₄): n_max3600–3300 (OH, broad), 1760 (CvO), 1380,

1370 (dioxolane absorption) cm²¹. ¹H NMR: complex due

to the presence of two isomers. This mixture was used as

such in the next step.

5.1.31. (3R/S,5R)-5-[10-(Tetrahydropyran-2-yloxy)-

decyl]-3-{[(4R)-2,2-dimethyl-[1,3]dioxolan-4-yl]-

hydroxy-methyl}-dihydrofuran-2-one 39. To a solution of

di-isopropylamine (450 mg, 4.5 mmol) in THF (20 mL) at

0 8C a solution of BuLi (2.5 mL, 1.6 N) in hexane was

added. The stirred solution was cooled to 270 8C and

lactone 37 (1.1 g, 3.4 mmol) dissolved in THF (10 mL) was

added drop by drop. After stirring for 15 min excess (4S)-

2,2-dimethyl-[1,3]-dioxolane-4-carbaldehyde 38 (1.17 g,

9.0 mmol) in THF (5 mL) was added in one portion,

followed by HMPA (1 mL). The reaction mixture was left

for 16 h at 270 8C and was quenched with satd aq. NH₄Cl

solution. The mixture was extracted with diethyl ether, the

organic phase was washed with water and dried over

MgSO₄. The residue obtained after work-up was chromato-

graphed over silica gel (100 g, EtOAc–hexane 3:1,

followed by 2:1). This gave starting material (0.41 g,

40.8%) and the desired product 39 (400 mg, 26%) as a

mixture of isomers which was processed as such in the next

step. IR (neat): n_max3600, 3480 (OH), 2980, 2920, 2840

aliphatic CH₂, 1765 (CvO), 1380,1370 (dioxolane

5.1.28. (5R)-5-Decyl-3-[(4R)-2,2-dimethyl-[1,3]dioxolan-

4-ylmethylene]-dihydrofuran-2-one 34. To a stirred

solution of the isomeric mixture 33 (800 mg, 2.27 mmol)

in dichloromethane (20 mL) and pyridine (10 mL) kept at

absorptions) cm²¹

.

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

5251

5.1.32. (5R)-5-[10-(Tetrahydropyran-2-yloxy)-decyl]-3-

{[(4R)-2,2-dimethyl-[1,3]-dioxolan-4-yl]-propylidene}-

dihydro-furan-2-one 40. The mixture of alcohols 39

(400 mg, 0.88 mmol) was dissolved in CCl₄(10 mL) and

methanesulfonyl chloride (120 mg, 1.04 mmol) and Et₃N

(120 mg, 1.2 mmol) were added. The reaction was followed

by IR. The alcohol absorption at 3600 and 3480 cm²¹

disappeared and strong absorptions at 1380 and 1180 cm²¹

from the sulfonate were emerging. In addition the CvO

absorption shifted to 1775 cm²¹. The solution was washed

with aqueous bicarbonate solution and dried (MgSO₄). After

evaporation of the solvent, the crude oily mesylate was

immediately used in the next step.

O), 2.86 (1H, m, CH–CvO), 2.54 (1H, m, CHH–CH–

CvO, lactone ring), 2.20 (1H, m, CHH–CH–CvO, side

chain), 1.58 (1H, m, CHH–CH–CvO, lactone ring), 1.54–

1.20 (25H, m, including dioxolane 2£Me at 1.41 and 1.35)

ppm. ¹³C NMR: 178.62 (CvO), 109.20 (O–C–O), 79.11

(CH–O–CvO), 73.49 (OCH₂–CH–O), 69.23 (O–CH₂–

CH–O), 63.07 (CH₂–OH), 38.24 (CH₂–CvO), 35.52

(CH₂–CH–CvO, lactone ring), 34.20 (CH₂–CH–CvO,

side chain), 26.96 (C–CH₃), 25.65 (C–CH₃), other CH₂

signals at 35.43, 32.80, 29.50, 29.39, 29.38, 29.30, 25.71,

and 25.21 ppm. Another CH₂seems to be hidden at

29.28 ppm.

5.1.36. (5R,3S,4⁰R)-5-Dec-(9⁰⁰)-enyl-3-(2⁰,2⁰-dimethyl-

[1⁰,3⁰]dioxolan-4⁰-yl)-methyl)-dihydrofuran-2-one 44. To

a solution of alcohol 43 (69 mg, 0.2 mmol) in dry THF

(5 mL) kept under argon, o-nitrophenylselanylcyanide

(68 mg, 0.3 mmol) was added, followed by tributyl-

phosphine (61 mg, 0.3 mmol) in a small amount of THF.

Soon a red colored solution appeared. The mixture was

stirred for 2 h at room temperature, after which the starting

material had disappeared. After evaporation the residue was

chromatographed over silica gel (diethyl ether–hexane 1:1).

This gave 98 mg (91%) of the selenide, which was

processed further immediately.

To a solution of the mesylate in diethyl ether (10 mL), DBU

(200 mg, 1.33 mmol) was added. The amine·HCl salt

formed immediately started to precipitate. According to

TLC two products had been formed (cis and trans alkene

40). After 2 h the reaction mixture was shaken with 10% aq.

tartrate solution and the ether phase was dried over MgSO₄.

After evaporation of the solvent the residue was chromato-

graphed over silica gel (diethyl ether–hexane 1:1). This

gave 40 as an oily mixture of isomers (247 mg, 72.5%) that

was used as such in the next step. IR (neat): 1755 (CvO),

1680 (CvC) cm²¹

.

5.1.33. (5R)-5-(10-Hydroxy-decyl)-3-[(2R)-2,3-di-

hydroxy-propylidene]-dihydrofuran-2-one 41. The

To a stirred solution of the selenide in THF (5 mL) an excess

of H₂O₂(0.3 mL, 30%) was added. Stirring was continued

for 2 h. TLC pointed out that the reaction was complete.

Satd aq. bicarbonate solution was added and the mixture

was extracted with hexane. The solution was chromato-

graphed over silica gel. Elution with petroleumether–

diethyl ether 1:1 gave a yellowishly colored product 44

(68 mg). Some selenide material was apparently sticking to

the product. IR (CCl₄): n_max2980, 2920, 2855 (aliphatic

C–H), 1775 (CvO), 1635 (CvC), and the sharp dioxolane

absorptions at 1370 and 1380 cm²¹. This product was

subjected to hydrolysis right away.

mixture of cis and trans alkene 40 was dissolved in

MeOH (10 mL), and a small amount of TsOH was added

with stirring. The protecting groups were removed fast. The

reaction was monitored by TLC. A small amount of aq.

bicarbonate solution was added and the mixture was

concentrated in vacuo. Water was added to the residue

and the mixture was extracted with chloroform (4£). After

drying and evaporation of the solvent, triol 41 was obtained

as a solid (140 mg, 70%). The solid still showed a long

melting point range due to the presence of two isomers. IR

(KBr): n_max3500–3100 (OH), 1750 (CvO), 1680 (CvC)

cm²¹. This material was used as such in the next step.

5.1.37. (5R,3S,2⁰R)-5-Dec-9⁰⁰-enyl-3-(2⁰,3⁰-dihydroxy-pro-

pyl)-dihydrofuran-2-one 1 (rubrenolide). Compound 44

(crude 68 mg, 0.2 mmol) was dissolved in MeOH (5 mL). A

small amount of TsOH was addded and upon stirring a fast

deprotection of the diol function occurred (TLC). Aqueous

bicarbonate solution was added and the mixture was

concentrated in vacuo. The residue was extracted with

ether. The ether solution was dried over MgSO₄and the

solvent was removed after ﬁltration. The residue was

chromatographed over silica gel. Elution with hexane–

EtOAc removed the yellow colored by-product. Elution

with EtOAc gave pure compound 1 as a colorless solid

(36 mg, 60%), mp 99–100 8C, [a]²_D⁰¼þ21.2 (CH₂Cl₂,

c¼0.3), lit.³mp 100 8C, [a]_D²⁰¼þ21 (CHCl₃). The

5.1.34. (5R,3S,2⁰R)-5-(10-Hydroxy-decyl)-3-(2⁰,3⁰-di-

hydroxy-propyl)-dihydrofuran-2-one 42. Product 41 was

hydrogenated quantitatively in MeOH with H₂Pd/C to give

compound 42 as a solid. IR (KBr): n_max3500–3200 (OH),

2920, 2840 (aliphatic CH), 1740 (CvO) cm²¹. This

product was used as such in the next step.

5.1.35. (5R,3S,4⁰R)-5-(10-Hydroxy-decyl)-3-[(2⁰,2⁰-

dimethyl-[1⁰,3⁰]dioxolan-4⁰-yl)-methyl])-dihydrofuran-2-

one 43. A solution of the above mentioned triol 42 in

acetone was stirred with a small amount of TsOH. After

30 min aq. bicarbonate solution was added. The reaction

mixture was concentrated and the residue was extracted

with ether. After work-up the residue was chromatographed

over silica gel (diethyl ether–hexane 1:1). Product 43 was

obtained as a solid (115 mg, 73%), mp 78–79 8C (hexane).

Elemental analysis, calcd for C₂₀H₃₆O₅(356.477) C: 67.41,

H: 10.11%; found: C 67.31, H 9.94%. IR (KBr): n_max3500–

1

400 MHz H NMR spectrum was in complete agreement

with a 400 MHz NMR spectrum that was taken from the

authentic product.

5.2. Separation of rubrenolide 1 and rubrynolide 2

3300 (OH), 1770 (CvO), cm²¹

.

¹H NMR (400 MHz),

Professor Ollis (University of Shefﬁeld) kindly provided a

mixture of the natural product (78 mg) as it was isolated

from the natural source; 25 mg of this mixture was dissolved

in EtOH (0.5 mL, 95%) and 5% AgNO₃solution in EtOH

(CDCl₃): d 4,35 (1H, m, CH–O–CvO), 4.13 (1H, m,

CH₂–CH–O–CMe₂), 4.07 (1H, m, O–CH–CHH–O), 3.64

(2H, t, J¼6.6 Hz, CH₂–OH), 3.57 (1H, m, O–CH–CHH–

5252

L. Thijs, B. Zwanenburg / Tetrahedron 60 (2004) 5237–5252

14. Marshall, J. A.; Flynn, G. A. J. Org. Chem. 1981, 44, 1391. For

the preparation of the selenide see: Grieco, P. A.; Gilman, S.;

Nishizawa, M. J. Org. Chem. 1976, 41, 1485.

(95%, 1 mL) was added. A white precipitate formed

immediately. The solid was ﬁltered off and washed with

EtOH. The ﬁltrate was concentrated and the residue was

treated with water. The mixture was extracted with ether.

Drying over MgSO₄, ﬁltration and evaporation of the

solvent gave rubrenolide (7 mg, mp 98 8C), which was used

as reference material. The solid that was ﬁltered off, was

treated with a 10% NaCN aq. solution and extracted with

ether. After work-up 16 mg of rubrynolide was obtained.

This compound was hydrogenated in EtOH with Pd/C as

catalyst to give dihydrorubrenolide 36 which was used as

reference material (for the 400 MHz spectrum, see

preparation of epi-dihydrorubrenolide 28).

15. Jackson, D. Y. Synth. Commun. 1988, 18, 337.

16. Rasmussen, J. R.; Slinger, C. J.; Kordish, R. J.; Newman-

Evans, D. D. J. Org. Chem. 1981, 46, 4843. See also:

Rasmussen, J. R. J. Org. Chem. 1980, 45, 2725.

17. For a related structure see: Tamamura, H.; Bienfait, B.; Nacro,

K.; Lewin, N. E.; Blumberg, P. M.; Marquez, V. E. J. Med.

Chem. 2000, 43, 3209.

18. Doolittle, R. E.; Tumlinson, J. H.; Proveaux, A. T.; Heath,

R. R. J. Chem. Ecol. 1980, 6, 473. The assignment R and S

were made opposite to that of the correct structure.

19. An NMR of 31 in the presence of the chiral shift reagent

Eur(hfc)₃indicated an ee of 91%.

20. Hubschwerlen, C. Synthesis 1986, 962.

21. Klyne, W.; Scopes, P. M.; Wlliams, A. J. Chem. Soc. 1965,

7237.

References and notes

1. Franca, N. C.; Gottlieb, O. R.; Coxon, D. T.; Ollis, W. D.

J. Chem. Soc., Chem. Commun. 1972, 514.

22. The late Professor Ollis, kindly provided us with a free sample

of the natural product consisting of a mixture of rubrenolide

and rubrynolide. The two products could be separated as

reported.³Rubrynolide was hydrogenated to give dihydro-

rubrenolide (see Section 5).

2. Gottlieb, O. R. Phytochemistry 1972, 11, 1537. Filho, R. B.;

Diaz, P. P.; Gottlieb, O. R. Phytochemistry 1980, 19(1980),

455.

3. Franca, N. C.; Gottlieb, O. R.; Coxon, D. T. Phytochemistry

1977, 16, 257.

23. Horeau, A.; Nouaille, A. Tetrahedron Lett. 1971, 1939.

24. Horeau, A.; Kagan, H. B. Tetrahedron 1964, 20, 2431.

´

25. Cherest, M.; Felkin, H. Tetrahedron Lett. 1968, 2205. Anh,

4. Cahn, S. R.; Ingold, C.; Prelog, V. Angew. Chem. 1966, 78,

413.

N. T.; Eisenstein, O. Nouv. J. Chim. 1977, 1, 61.

26. Heathcock, C. H.; Young, S. D.; Hagen, J. P.; Pirrung, M. C.;

White, C. T.; Van Derveer, D. J. Org. Chem. 1980, 45, 3846.

27. Jurczak, J.; Pikul, S.; Bauer, T. Tetrahedron 1986, 42, 447, and

references cited herein.

5. Hussain, S. A. M. T.; Ollis, W. D.; Smith, C.; Fraser Stoddart,

J. J. Chem. Soc., Perkin Trans. 1 1975, 1480.

6. Taylor, S. K.; Hopkins, J. A.; Spangenberg, K. A.; McMillen,

D. W.; Grutzner, J. B. J. Org. Chem. 1991, 56, 5951.

7. Saito, T.; Thijs, L.; Ettema, G.-J.; Zwanenburg, B.

Tetrahedron Lett. 1993, 34, 3589.

28. Minimum energy structures were generated by semi-empirical

MOPAC PM3 calculations and carried out by Dr G.J.A.

Ariaans (Synthon, Nijmegen, The Netherlands).

8. For other examples see: (a): Waanders, P. P.; Thijs, L.;

Zwanenburg, B. Tetrahedron Lett. 1987, 28, 2409.

(b) Dommerholt, F. J.; Thijs, L.; Zwanenburg, B. Tetrahedron

Lett. 1991, 32, 1495. (c) Dommerholt, F. J.; Thijs, L.;

Zwanenburg, B. Tetrahedron Lett. 1991, 32, 1499.

9. Thijs, L.; Dommerholt, F. J.; Leemhuis, F. M. C.;

Zwanenburg, B. Tetrahedron Lett. 1990, 31, 6589.

10. Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102,

5974.

29. There are examples of complete control of stereochemistry

when both components in the Aldol condensation are chiral,

´

see: Dongala, E. B.; Dull, D. L.; Mioskowski, C.; Solladie, G.

Tetrahedron Lett. 1973, 4983. see also: Masamune, S.; Choy,

W.; Petersen, J. S.; Sita, L. R. Angew. Chem., Int. Ed. Engl.

1985, 24, 1.

30. Girlanda-Junges, C.; Keyling-Bilger, F.; Schmitt, G.; Luu, B.

Tetrahedron 1998, 54, 7735.

11. Bal, B. S.; Childers, W. E.; Pinnick, H. W. Tetrahedron 1981,

37, 2091.

31. Ermilova, E. V.; Remizova, L. A.; Favorskaya, I. A.;

Tregubova, N. L. Zh. Org. Khim. 1975, 11, 520.

32. Kang, S. K.; Kim, Y. S.; Lim, J. S.; Kim, K. S.; Kim, S. G.

Tetrahedron Lett. 1991, 32, 363.

12. Thijs, L.; Waanders, P. P.; Stokkingreef, E. H. M.;

Zwanenburg, B. Recl. Trav. Chim. Pays-Bas 1986, 105, 332.

13. Russell, T. W.; Hoy, R. C. J. Org. Chem. 1971, 36, 2018.

Article Doi

Rubrenolide, total synthesis and revision of its reported stereochemical structure

DOI: 10.1016/j.tet.2004.04.037

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.tet.2004.04.037

Full text of DOI:10.1016/j.tet.2004.04.037

Products guided by the article

R&D Labs maybe for 725225-78-5

Relevant to this article

Hot Product