SYNTHESIS OF 8-METHYLNONANE-1,6,7-TRIEN-4-ONE AND RELATED ALLENES AS POTENTIALLY USEFUL SYNTHETIC PRECURSORS

The allenic ketone 8-methylnonane-1,6,7-trien-4-one and related allenes have been synthesized from simple commercially available materials. Since allenes analogous to 8-methylnonane-1,6,7-trien-4-one have previously been transformed to substituted bicyclo[3.3.0]octanones via corresponding bicyclo[3.2.0]heptanones, it is anticipated that the present allenic ketone may also undergo similar transformations. Substituted bicyclo[3.3.0]octanones are known synthetic precursors of tricyclic sesquiterpenes. Thus, 8-methylnonane-1,6,7trien-4-one presents itself as a possible precursor for the synthesis of tricyclopentanoid ring system present in sesquiterpenes such as hirsutene and ∆-capnellene.


RESULTS AND DISCUSSION
Initially we designed a synthetic strategy that aimed at 5-methyl-3,4-hexadienal ( 16) as a key intermediate towards the targeted allenic ketones 4a and 4b.One obvious way to obtain compound 16 would be to oxidize its alcohol precursor, that is 5-methyl-3,4-hexadienol (17).Attempts were made to prepare the dienol 17 by first employing a Claisen type [3,3] sigmatropic rearrangement of the propargyl vinyl ether 18, which was generated in situ from 2-methyl-3butyn-2-ol, triethyl orthoacetate and propionic acid (Scheme 4), to give ethyl 5-methyl-3,4hexadienoate (19).On encountering a problem in the reduction of the β-allenic ester 19 to obtain the alcohol 16, a new route to the latter compound was conceived.In this case, 2-(2,2dibromo-3,3-dimethylcyclopropyl) ethanol (20) was envisioned as the immediate precursor to the allenic alcohol 17.Compound 20 was obtained in 84-87% yield from 21 through a Baeyer-Villiger oxidation followed by a base catalysed hydrolysis of the resulting ester 22.The ketone 21 was prepared in 90% overall yield from the readily available 6-methyl-5-hepten-2-one in three simple steps, which are: protection of the carbonyl group as a cyclic ketal [12], addition of the dibromo carbene and deprotection of carbonyl group in that order.Therefore, the overall yield of alcohol 20 from 6-methyl-5-hepten-2-one was 76%.
Having obtained compound 20 in good yield, the stage was set for the preparation of 5methyl-3,4-hexadienol (17), the immediate precursor for 16.The conversion of 20 to 17 was achieved in 95% yield by treatment of an ice-cooled ethereal solution of compound 20 with an ethereal solution of 1.5 M MeLi in an inert atmosphere.It is worth mentioning that 5-methyl-3,4-hexadienol (17) was prepared in 76% overall yield following a six steps procedure from 1,3propanediol [10].The preparation reported in the present work afforded the dienol 17 in comparable yields (72%) in six steps utilizing simple reagents.However, oxidation of the allenic alcohol 17 by the chromium trioxide-pyridine complex gave the isomerized aldehyde 5methyl-2,4-hexadienal (23) instead of the targeted aldehyde 16.The 2,4-hexadienal 23 was characterized on the basis of spectroscopic data.The appearance of the carbonyl group absorption at 1680 cm -1 in the IR spectrum is characteristic of α,βunsaturated aldehydes.The 1 H NMR spectrum indicated the presence of three olefinic protons (contrary to one olefinic proton as it would be for compound 16), a multiplet at 6.0 ppm due to the αH and the γH, and a dublet dublets at 7.5 ppm due to the βH.The appearance of a dublet for the adehydic proton at 9.5 ppm suggests the presence of one proton at the α-carbon (contrary to two protons as it would be the case of the allenic aldehyde 16).The singlet at 1.9 ppm is due to the gem-dimethyl protons.The UV spectrum showed an absorption maximum at 274.5 nm (ε = 17600) which is expected for a dienal chromophore.
Dissatisfied with this isomerization of the anticipated aldehyde 16 to 23, we terminated efforts to prepare 16 as a key intermediate towards the targeted allenic ketones 4a and 4b.Our second route towards the allenic ketone 4a is summarized in Scheme 5.In this strategy we aimed at the 1,3-dthiane derivative 24, which we anticipated would easily be converted to 4a. 2-Allyl-1,3-diathiane (25) was obtained in 90% yield from the lithium derivative of 1,3-dithiane (26) and allyl bromide.However, the conversion of 25 to 27 proved to be problematic, presumably because the intermediate, lithiated 25, could have reacted with the acetylenic proton of the propargylic bromide and thereby regenerating 25.To circumvent this problem, we decided to form the allene functionality prior to coupling with 25.We, thus, prepared the allenic bromopentadiene 29, albeit in very low yield, starting from the commercially available propargyl alcohol.The diol 28 was prepared by reacting acetone and the ethynyl Grignard reagent derived from propargyl alcohol.Treatment of 28 with HCl/CaCl 2 afforded 4-chloro-4methyl-2-pentyn-1-ol, which was converted, without isolation, to 4-methyl-2,3-pentadien-1-ol by reaction with LiAlH 4 .Treatment of the latter allenic pentadienol with PBr 3 furnished the allenic bromopentadiene 29, which was coupled with 25 to furnish 24 in 58% yield.The allene 24 was characterized on the on the basis of spectral data.The presence of the allenic bond is shown by an infrared band at 1955 cm -1 .The 1 H NMR spectrum is consistent with the structure.However the low yield of 29 coupled with the moderate yield of 24 left us with insufficient material for subsequent transformations.This was certainly not satisfactory, and we looked for alternative ways of synthesizing compound 4a.The fact that 2-(2,2-dibromo-3,3-dimethylcyclopropyl) ethanol (20) was obtained in 76% overall yield from 6-methyl-5-hepten-2-one prompted us to revisit our first strategy (Scheme 4, vide supra) with some modifications in mind.Thus, in our third strategy, we planned to form the aldehyde functionality prior to the introduction of the allenic moiety.This strategy aimed at circumventing the isomerization that occurred for the allenic aldehyde 16.This route towards the allenic ketone 4a is depicted in Scheme 6 [steps (i)-(iii) are included for completeness as they have already been discussed in Scheme 4].
Swern [13] oxidation of alcohol 20 furnished 2-(2,2-dibromo-3,3-dimethylcyclopropyl) ethanal (30), which was transformed, without isolation, to 1-(2,2-dibromo-3,3dimethylcyclopropyl)-4-penten-2-ol (31) in 71% overall yield from 20. Treatment of 31 with N-O-bis-trimethylsilylacetamide in the presence of a catalytic amount of trimethylchlorosilane [14] gave the silyl ether 32 in 90% yield.The silyl ether was converted to the vinyl allene 33 (61% yield) by treatment of 32 with MeLi at -5 °C, and Jones oxidation of the allene gave the desired ketone 4a in 61% yield.Infrared absorption at 1965 and 1705 cm -1 are characteristic of the allenic linkage and isolated carbonyl group, respectively, and the NMR spectra fully confirm the structure of the desired allenic ketone 4a.Scheme 6 Although we are not fully satisfied with the length (large number of steps) and linearity of the strategy, we nevertheless have successfully prepared the desired ketone 4a in 21% overall yield from compound 21.It is, therefore, possible to prepare sufficient amounts of compound 4a by scaling up the synthesis or conducting repeated preparations.We also have prepared the thioketal-protected target ketone in the form of compound 24 (Scheme 5).It is, therefore, worthy revisiting this shorter and convergent route towards 4a with the purpose to improve the yield of the bromopentadiene 29 and, subsequently, that of compound 24.Furthermore, it may be desirable to design a shorter and convergent route to 8-Methylnonane-1,6,7-trien-5-one, a regioisomer of 4a, using some of the compounds prepared in this work, for example, 4-methyl-2,3-pentadien-1-ol, which was the precursor for the bromopentadiene 29 in Scheme 5.

General
Reagents.Unless otherwise indicated all analytical grade reagents were used without further purification.Other grades were purified according to known procedures before use.
Instruments.Capillary gas-liquid chromatography (GLC) analyses were carried out on a Varian 3400 Gas Chromatograph using a supelcowax 10 wall-coated column of 30 m length and 0.25 mm internal diameter.Other analytical GLC were carried out on either a Hewlett Packard 5710a Gas Chromatograph using 3% SP 2100 packed column of 2.5 m length and 2 mm internal diameter or on Chrompack CP 9001 Gas Chromatograph using 3% OV-17 packed column of 2 m x ¼ " x 2 mm dimensions.Thin-layer chromatography (TLC) was carried out on aluminium plates pre-coated with silica gel 60 F 254 .The spots were visualized by illuminating with UV lump (254 nm) or by using p-anisaldehyde spray reagent.Infrared (IR) spectra were recorded on either a Perkin-Elmer Infrared Spectrometer 1310 or a Magna-IR Spectrometer 550.Absorptions are indicated as (s) = strong, (m) = medium, and (w) = weak.NMR Spectra: 1 H and 13 C NMR spectra were recorded on the following instruments: Jeol PMX60 SI NMR Spectrometer operating at 60 MHz, Varian Gemini-200 NMR spectrometer, varian XL-300 NMR Spectrometer, and dpx-300 NMR Spectrometer.Tetramethylsilane (TMS) was used as an internal standard.Ultraviolet (UV) spectra were recorded on a Shimadzu UV-VIS Recording Spectrophotometer UV-260.Melting points were taken on a Buchi melting point apparatus and are uncorrected.

2 Protonation of 6
Scheme 1It was anticipated that the [2+2] ene-allene cycloaddition of compound 4 should lead to the bicyclo[3.2.0]heptanone 6 and the acid catalysed rearrangement of 6 should furnish the bicyclo[3.3.0]octanone 5.The difference in the position of the gem-dimethyl substructure in compounds 5a and 5b can be mechanistically rationalized by the anticipated acid catalysed rearrangement of compound(s) 6 as depicted in Scheme 2.