Reduction lactol, borohydride

Although the nature of the general polar effect suggested by Kamernitzsky and Akhrem " to account for axial attack in unhindered ketones is not clear, several groups have reported electrostatic interactions affect the course of borohydride reductions. Thus the keto acid (5a) is not reduced by boro-hydride but its ester (5b) is reduced rapidly further, the reduction of the ester (6b) takes place much more rapidly than that of the acid (6a). Spectroscopic data eliminate the possibility that in (5a) there is an interaction between the acid and ketone groups (e.g. formation of a lactol). The results have been attributed to a direct repulsion by the carboxylate ion as the borohydride ion approaches. " By contrast, House and co-workers observed no electrostatic effect on the stereochemistry of reduction of the keto acid (7). However, in this compound the acid group may occupy conformations in which it does not shield the ketone. Henbest reported that substituting chlorine... 

Conjugate addition of methyl magnesium iodide in the presence of cuprous chloride to the enone (91) leads to the la-methyl product mesterolone (92) Although this is the thermodynamically unfavored axially disposed product, no possibility for isomerization exists in this case, since the ketone is once removed from this center. In an interesting synthesis of an oxa steroid, the enone (91) is first oxidized with lead tetraacetate the carbon at the 2 position is lost, affording the acid aldehyde. Reduction of this intermediate, also shown in the lactol form, with sodium borohydride affords the steroid lactone oxandrolone... 

The neutralization values were influenced by reduction with strong reducing agents, lithium aluminum hydride, sodium borohydride, and amalgamated zinc plus hydrochloric acid (35, 46). For the most part, the consumption of NajCOj and of NaOEt decreased in equivalent amounts. This is further confirmation of the assumption that lactones of the fluorescein type or of the lactol type are present. The reaction with sodium ethoxide was shown to be no true neutralization, that is, exchange of H+for Na+, at all, but an addition reaction w ith the formation of the sodium salt of a semi-acetal or ketal ... 

Tetrahydrothiazin-3-ones are lactams that have been reduced to tetrahydrothiazines with borane <1980JHC449>, sodium borohydride <1992JOC4215>, or LAH <1987H(26)1503>, without cleavage of carbon-sulfur bond. In one case, incomplete reduction occured with LAH the intermediate lactol was dehydrated to give a dihydrothiazine as main product <1989JPS937>. 

Based on the precedented reduction of 2//-dihydropyrones,86 the combination of Lewis acid and hydride source exemplified by Et3SiH/BF3 seemed ideally suited to our needs (Scheme 9). While (+)-artemisinin 1 could not be reduced directly to 10-deoxoartemisinin 108 with Et3SiH/BF3, dihydroartemisinin (175, R = H) was smoothly converted at low temperature to desired tetrahydropyran 108 in 96% yield. Further, this method was insensitive to scale being readily accomplished on the gram or milligram level. It was also found that small scale reductions could be more conveniently conducted utilizing diisobutylaluminum hydride in place of sodium borohydride. As applied to the problematical case, it was found that lactone 125 could be reduced to lactol and thence 115 as outlined in Scheme 3 in excellent yield. Furthermore, the yield for the conversion of lactone 45 into 9-butyl-10-deoxoartemisinin 112 could be similarly improved from 58 to 90%. 

Methanolysis of a C-20 mixed anhydride of gibberellin A13 gave the unusual 19-ortho-ester, the structure (71) of which was established by X-ray analysis.117 The formation of this compound and the corresponding 19-epimeric 20->19-lactols by sodium borohydride reduction of the mixed anhydrides reveals the facile participation of the 19-esters in the reactions of C-20. This feature may be of biosynthetic significance. 

On the other hand, the xylofuranoside 28 was benzylated, followed by removal of the isopropylidene group and subsequent benzylation to give the tribenzylated derivative 31. Acid hydrolysis of 31 followed by sodium borohydride reduction of the resulting lactol and subsequent mesylation and then selective nucleophilic displacement of the primary mesylate by sodium azide in DMF afforded the azido mesylate 32. Reduction of the azide was accompanied by cyclization and deprotection to afford 2 in 11% overall yield from 28. 

Tin(ll) chloride reduction of the azidotriflates 81 followed by intramolecular cycliza-tion with sodium acetate in methanol and subsequent protection of the resulting secondary amine with benzyl chloroformate afforded the a- and -furanoside carbamates 83 and 82. Hydrolysis of 82 and 83 by TEA in aqueous dioxane followed by sodium borohydride reduction of the resulting lactol furnished the protected 1-deoxynojirimcin 84 (49%). Subsequent hydrogenation of 84 afforded 2. 

The peroxide bond in artemisinin (9) resists reduction with sodium borohydride, yielding a mixture of the epimeric lactols or dihydroartemisinins (65) <82MI 620-0l>. On the other hand, hydrogenation of (9) over Pd/CaCOs converts it into deoxyartemisinin (19), presumably by spontaneous dehydration of the intermediate diol (94) (Scheme 11) <88JMC645>. 

A literature survey revealed a single report dealing with the synthesis of this class of C-nucleosides (74JOC1374). Reaction of 3-benzyl-4-lithio-l,2,3-oxadiazol-5-one with the L-gulonolactone derivative 714 stereospe-cifically afforded the l,2,3-oxadiazol-4-yl C-nucleoside lactol 715. Reduction of 715 with sodium borohydride gave the acyclo C-nucleoside 716 (Scheme 192). 

C-1 Selective reduction of malates is not restricted exclusively to the diesters, but succeeds with anhydrides as well. Selective reduction of anhydride 54 at the C-1 site with sodium borohydride affords lactone 55 in 61% overall yield from 7b. A second reduction of the lactone carbonyl with diisobutylaluminum hydride furnishes lactol 56, which is then converted to acetal 57 with 2,2-dimethylpropane-l,3-diol. Introduction of the required acetylene group requires an additional 5 steps. 

An enantioselective synthesis of ( —)-(2i ,3i ,6S,8i )-epi-nonactic acid (124) is shown in Scheme 89. The reduction of (122) with diethylmethoxyborane/sodium borohydride yields a iyn-diol (not isolated), which cyclizes spontaneously to the lactol (123). Pyridinium -toluenesulfonate-catalyzed dehydration, desulfurization with Raney nickel, Rh/Al203-catalyzed hydrogenation, and hydrolysis of the ester group affords (124) <94JOC3898>. 

Indeed, Balsevich and Bishop reported difficulties while attempting to reduce nepetalactone (40) with sodium borohydride. The desired lactol 66 was not isolated, with the saturated lactone 67 and diol 68 observed (Scheme 15). The authors proposed that both products resulted from methano-lysis of the lactone ring to provide dicarbonyl 69, with subsequent reduction affording a mixture of the two products.It can be envisioned, however, that diol 68 could form via reduction of lactone 40 to lactol 66, followed by ringopening and subsequent reduction of the hfs-aldehyde 70 over the course of the 20 h reaction. 

---