Easy borohydride and CTH reduction of nitrostyrenes

Abstract

Using a two step reduction, which requires very forgiving reaction conditions it is possible to reduce nitrostyrenes (nitrostyrenes also include nitropropenes henceforth in this document) to alkylamines in very high yield. The two steps are performed without isolation or purification of the intermediate products, and even uses the same solvent throughout the reactions.

The first step is a borohydride reduction of the double bond, followed by a CTH reduction using Pd/C and potassium formate, of the nitro group to form the desired amine. The potassium formate can be prepared from potassium hydroxide and formic acid immediately before use, without any need of isolation.

GC-MS analysis of the final amines has shown that a distillation of the final product is not required, as the purity has been 98%+ in all non-distilled trials. The GC-MS apparatus was not accurate enough to tell a distilled sample from a non distilled sample, which indicates a very pure product.

Chemistry

Reducing the double bond with sodium borohydride in IPA and water, will give the nitroalkane in high yields. The IPA layer isolated from the reduction is topped up with 20% more IPA and used directly in the next step.

The IPA containing the nitro compound is simply mixed with the 5% Pd/C, and a premade slurry of KOH and HCOOH is added. The whole mess is stirred at 75-80^oC for a few hours, whereafter the reaction is over. The nitrogroup is reduced with potassium formate as the hydrogen donor. Potassium gives much superior yields to ammonium and sodium. The Pd acts as a hydrogenation transfer catalyst. The amount of water present is very critical for both the yield and the swiftness of the reaction. A 2.5-3 molar equalivent of water to formate usually gives the highest yields. The reaction rapidly and is usually over in 1-3 hours. Once hydrogen evolution ceases the reaction can be considered done.

Experimental

The Reduction

This example uses (p-fluoro-phenyl)-2-nitropropane as the substrate. This can of course be substituted with any other similar nitroalkane and applied in the same manner.

The following amounts where used. The amount of the nitropropane can simply be switched with the amount of nitropropene used, as the first reduction is so high yielding.

50 mmol nitroalkane
250 mmol potassium formate¹
750 mmol water
25-50 ml IPA (depending on how much is needed to get a nice solution)²
10%w/w 5%Pd/C (catalyst to substrate)

The IPA layer from the previous reaction was topped of with +20% more to make it a little more dilute. 0.7 g of 5% Pd/C was added to the IPA layer and stirred heavily.

In another beaker 7 mL's of water was made into a slurry with 14 g's of dry KOH. While stirring like a madman, 13.5 g 85% HCOOH was dripped in (caution: very exothermic). Do the math yourself on how much water to add. You get water from: the neutralization of the two acids, the water in the formic acid, the KOH and perhaps the Pd/C if you use prewetted.

The potassiumformate mixture was poured into the nitro/IPA flask in one portion. The whole lot is heated to 75^oC while stirring. Evolution of hydrogen was noted around 50^oC. A few mL's of 80% acetic acid was added now and then, when the mixture got too "lumpy" (4 mLs was used in total). After two hours the hydrogen evolution had subsided and the reaction was deemed over.

Workup

The black slurry was filtered twice through filter paper (first time alot of black sludge came through, but simply use the same filter as before as a "plug" for the second filtering). The filtercake was rinsed with a little IPA to get all the goodies with it. The final filtrate was almost colourless. After saving the filtrate, the filtercake was washed with water, some dilute hydrochloric acid, and finally some pure water to remove acid traced. This can then be scraped up and saved in the "used 5% Pd/C pile".

The almost-clear filtrate was saturated with table salt and filtered (the Pd/C filtercake can actually be re-used here and then cleared up with water). The lower aqeous layer was yet again discarded and the IPA stripped at atmospheric pressure.

The yellowish oil residue weigthed 6.5 grams. This was dissolved in dilute hydrochloric acid and washed twice with 15 mL's DCM, giving a nearly colourless solution (perhaps a tint of green/yellow?). The aqeous layer was basified and extracted with 3x20 mL DCM. The DCM was dried with a tiny bit of MgSO₄, filtered and stripped. This left 4.5 grammes of a clear oil (p-F-amphetamine). The yield the total reduction is 72% with very easy steps in between .

The p-F-amphetamine was dissolved in IPA, and precipitated with a precisely half molar amount of H₂SO₄ dissolved in IPA. This was added in a thin stream and when the whole mess was unstirrable it was filtered and the filtrate titrated with the acid again. Yield of the sulfate salt was quantitative (95%+), which is good enough for me ! GC-MS analysis showed 98%+ purity of the compound, which is as high as that particular CG-MS goes. Thus no distilliation of the amine is nescassary.

Conclusion:

A similar reduction of (2,5-dimethoxy-phenyl)-2-nitropropene gave an overall yield of 86% in the two step reduction. The method followed was exactly the same as described above

With a little practice the reduction from the nitropropene to the amphetamine can be done in about five hours total. The yields are really good, and reaction conditions are mild. It does not seem to bother the reactions, that the same IPA is used throughout the scheme. So please do reduce away :-)

Pretty slick dare i say!