# How to add the two hydroxy groups to my raw material with as high as possible yield?

From the raw material (above), how to add the two hydroxy groups in red(below)? Is there any proven successful synthesis route with higher yield (40% or above)?

-
I would consider synthesis from acetophenone. Condensation of acetophenone enolate-anion with phosgene should give triketone, which in turn may be reduced by proper reagent into target product. Methods of introduction of substitute into benzilic position do exist, and some of them will work here, but I doubt you'll find anything working in a way you wish. –  permeakra Aug 4 '14 at 14:21
I've given an answer. That said, I think @permeakra is definitely suggesting a vastly superior approach, assuming you're not obligated to use the 1,5-diphenyl-3-pentanol as a starting material. –  Greg E. Aug 4 '14 at 14:33
Two questions, 1) do you really need to start with the mono alcohol pictured above? 2) Your product contains 3 chiral centers resulting in the possible production of 4 diastereomers (2 chiral, 2 meso), is there one in particular you are interested in, or is a mixture acceptable? –  ron Aug 7 '14 at 15:53

A few possibilities for direct hydroxylation:

• $\ce{SeO2}$ immediately comes to mind as one potential option; it is well known to selectively oxidize allylic positions to yield alcohols and carbonyl compounds, and there appear to be references suggesting it reacts at the benzylic position as well.
• In Advanced Organic Chemistry, March notes that cerium(IV) triflate "converts benzylic arenes to benzylic alcohols, although the major product is the ketone when > 15% of water is present." Ref: Laali, K.K.; Herbert, M.; Cushnyr, B.; Bhatt, A.; Terrano, D. J. Chem. Soc., Perkin Trans. 1 2001, 578.
• Intriguingly, March also reports a variety of oxidations accomplished enzymatically. I have no experience and very limited knowledge of these reactions, so I'll just give the relevant primary literature references:
1. Adam, W.; Lukacs, Z.; Saha-Moller, C.R.; Weckerle, B.; Schreier, P. Eur. J. Org. Chem. 2000, 2923.
2. Hamada, H.; Tanaka, T.; Furuya, T.; Takahata, H.; Nemoto, H. Tetrahedron Lett. 2001, 42, 909.
3. Adam, W.; Lukacs, Z.; Harmsen, D.; Saha-Moller, C.R.; Schreier, P. J. Org. Chem. 2000, 65, 878.

If those approaches aren't appealing, there are various less direct routes. Radical bromination (with, e.g., NBS under photochemical conditions), followed either by direct substitution with hydroxide, or alternatively elimination followed by subsequent acid-catalyzed hydration, would likely work well. If you need stereoselectivity, there are are numerous epoxidation reactions (that is, after bromination and subsequent elimination) that would likely be appropriate (Sharpless, Jacobsen, Shi, and certainly numerous others). Conveniently, your molecule would already have the allylic alcohol necessary for Sharpless asymmetric epoxidation, should you want to go that route. In any case, it should be possible to open the epoxides at the less hindered carbon with a suitable hydride donor (e.g., LAH) to yield the alcohols after work-up.

-
most oxidants will attack the alcohol group, radical bromination is statistical and acid will likely dehydrate the target molecule, especially if partially oxidated, into conjugated system. Because of it, most simple solutions targeting benzyl position will not work desired way. –  permeakra Aug 4 '14 at 18:55
@permeakra, well, oxidation of the alcohol isn't necessarily fatal; oxidation to a ketone can be reversed by hydride reduction easily, and none of the oxidants I've mentioned are aggressive enough to cleave $\ce{C-C}$ bonds. I agree with you about bromination. Even with the preference for the benzylic carbon, there will probably be roughly equal amounts of 1,1- and 1,5-dibrominated products, and they would be difficult to separate. In any case, I agree that the process would be comparatively painful and inefficient. –  Greg E. Aug 4 '14 at 19:11

Actually, hell with that. I recall that organolithium compounds may be carefully oxidated at very low temperatures by different things, namely halogens, sulfur and - tadam! - molecular oxygen.

$\ce{BuLi}$ should selectively deprotonate each benzyl position once, and then the lithium compound may be very carefully oxidized with molecular oxygen at very low temperature followed by water hydrolysis into peroxide. Reduction of the peroxide should give the target compound, but it highly preferable to do it in basic conditions and avoid contact of target compound with acids as it looks to be able to dehydrate in acidic conditions. Another possible direction is to transform organolithium compound into boric acid and then oxidate it with hydrogen peroxide in basic conditions.

-
The idea of lithiation followed by oxidation to give the peroxide is really ingenious. It's not a reaction I see often, but this strikes me as an extremely elegant solution. Well done. –  Greg E. Aug 4 '14 at 19:39

It's worth trying Greg's suggestion to do benzylic oxidation or bromination/hydrolysis. If it works it's cool, but a lot of times trying to change 2 functionalities of a molecule in one go has poor results with poor yields and formation of mixtures of products. Additionally you might find that you will need to protect your alcohol first (that's more possible with permeakra's enolate oxidation approach).

Another approach could be to start from acetone or an acetone equivalent (like its silyl enol ether) and do a double condensation with benzaldehyde. I'm sure you can find similar reactions in the literature.

-