This file is a part of the Rhodium site archive. This Aug 2004 static snapshot is hosted by Erowid
as of May 2005 and is not being updated. > > Back to Rhodium Archive Index > >

Notes on Birch-like Reductions

by Chem Guy

Aurther Birch wrote about the reduction of benzyl alcohols in "The Journal of the Chemical Soceity" (1945 page 809) . His procedure was to have the benzyl alcohol dissolved in an amine-alcohol solution and then add the alkali metal. The molar product yeild was about 75% for carbinol Note: The ammonia was "about 30 times the volume of alcohol".

The Journal Of Organic Chemistry, Vol 40, 1975, page 3152 describes the birch method of benzyl alcohol reduction as so:

"SODIUM-AMMONIA-ETHANOL. To a stirred mixture of 771 mg (4.99 mmol) of tetralol and 500 mg (10.96 mmol) of EtOH in 20 mL of NH3 was added six peices of Na (253 mg, 11 mg-atoms) over a 14 minute period to maintain a dark blue solution. Approximately 12 minutes later the mixture turned white and then the NH3 was allowed to evaporate. Work up as described above yeilded a mixture of tetralin (75%) and 5,8-dihydrotetralin (25%)."

Now the same article in JOC lists a table of several benzyl alcohols and their product yeilds. All the fused cyclic alcohols and ring substitued benzyl alcohols have a low yeild with the Na-EtOH-NH3 method but the other single phenyl alcohols have a yeild around 100%. Never does the article refer to ephedrine but through analogy you can see that the Na-EtOH-NH3 method has its merits.


This is from "The Journal Of Organic Chemistry", Vol. 28, page 1094

"... In the course of this paper we are reporting the effects of water and alcohols on the course of the [Li-amine] reduction." "We found that, in the presense of alcohol, the Li-amine combonation can be used quite sucessfully to form dihydro aromatics." "The soduim-ammonia-alcohol system is not capable of reducing an unconjugated double bond except in certain isolated systems. The same is not true of the Li-amine system, which is capable of 1, 2 reductions of olfeins. Hence, with excess metal, the 2, 5 dihydro products can suffer a slow reduction of one double bond in the Li-amine system, thereby forminga 1-substituted olefin. In the Birch method excess metal (e.i., more than 2 equivalents based on the aromatic) can be used without fear of further reduction..." "Interestingly enough, the presense of 3 equivalents of ethanol was not capable of stopping the Li-amine reduction at the dihydro stage in the presence of 6 equivalents of metal." "Even 6 equivalents of ethanol did not prevent this isomersation entirly in the presence of 6 equivalents of Li, since 26% of the product in this case was 1-isopropylcyclohexene. Albeit, the major product was 2,5-dihydrocumene."[ 64% ] "The reaction in this case was quite exothermic and the metal was used up entirely in about 4 hours. Obviously much of the metal consumed was by direct reaction with the alcohol to form hydrogen. The results clearly indicate the necessity of controlling the quanity of Li used if one desires to prepare the unconjugated diene to the virtual exclusion of 1-substitued olefin by the Li-amine system." "A similar observation was noted when an attempt was made to reduce cumene in methylamine was water as the protn donator rather than an alcohol. In this case only 4 equivalents of metal were used, but still monoolefins were produced..." "One must conclude that the water was not able to prevent diene conjugation, ..." "It is at first sight startling that Li-amine reductions can be carried out successfully in the presence of water. This clearly indicates the extreme ease and rapidity with which electron transfer from the Li to the aromatic and its diene intermidiates occurs in this system. Obviously these organic species are competeing successfully for electrons with the hydrogen of the water molecules."

The reduction of isopropylbenzene. The birch reduction with sodium gets 92% 2,5 dihydroisopropylbenzene. The Li-amine gets 88% 2,5 dihydroisoproplylbenzene.

NOTE:This method (Li-amine) still can over reduce as show by this example, but it illustrates how sucessfully the aromatic competes for the electron with the water.

When the reduction of isopropylbenzene takes place with out water, with a trace of water, and with 2 eqivalents of water in MeNH2 and with 4 equivalents of Li.

W/O H2O-100% isopropylhexene

with a trace of H2O- 100% "

With 2 equivalents H2O- 93% "

Also this the results from using 6 equivalents of Li in the presence and absence of ethanol in EtNH2

Pure EtNH2- 71% isopropylhexene and 29% isopropylhexane

3 equivalents EtOH- 77% isopropylhexene and 23 isopropylhexane

6 equivalents of EtOH- 30% isopropylhexene, 2% isopropylhexane and 64% dihydro-isopropylbenzene.


From "Electrons in Liquid Ammonia" by J.C. Thompson, 1976

"The fact that dilute solutions containing equivalents amounts of alkali and alkaline metals give virtually the same near-infrared absorption spectra indictaes the presence of a common absorbing species which must be described without reference to the cation. Indeed, dilute solutions of solvated electrons electrochemically generated in the presence of tetraalkylammonium ions with widely-varying structural parameters are also optically indistinguishable from those formed by the dissolution of metal atoms."


From "The Journal of Chemical Physics", Vol 44, number 6, page 2297, 15 march 1966.

"...The approximate times necessary at -78C for the optical density due to the solvated electron to decay to one half its initial value were 8, 20, 8, and 25 mirco-sec for methanol, ethanol, isopropanol, and n-butanol, respectively. These half-times are 10 to 20 times longer than the corresponding values at room tempature."

"... at -78C, absorptions due to the solvated electron were obtained for monomethylamine (T[.5]= 3 mirco-sec) and monoethylamine (T[.5]= 3.5 mirco-sec). At -110C the absorption of the solvated electron in diethylether was obtained (T[.5]= 2 mirco-sec)."

Here is a condensed table of the half-life of the solvated electron, (T[.5]), in various solvents: (From the same work as above)

SUBSTANCE/HALF-LIFE IN MIRCO-SECONDS [all measuements at 25C unless otherwise stated] 100% glycerol / .44 micro-sec 63% glycerol- 37% water / .75 47% " - 53% " / .9 32% " - 68% " / 1.6 19% " - 1% " / 1.5 8.3% " - 92% " / 1.3 53% water - 47% ethanol / 3.4 36% " - 64% " / 2.2 20% " - 80% " / 2.5 10% " - 90% " / .40 50% ethylene glycol - 50% water / .85 @ 20C 10% " - 90% " / .6 @ 20C 70% methanol - 30% water / 2.7 79% isopropanol - 21% water / .7 @ 20C 70% " - 30% " / .65 @ 20C 31% glycerol - 69% ethanol / .5 12% " - 88% " / .75 50% methanol - 50% isopropanol / 10 @ -78C

From "Ionizing Solvents" by I. Junder 1970

from a table and its amendments concerning the solublities of compounds in liquid ammonia.

"NaOH....i[meaning insoluble in liquid ammonia] Na2SO4...i(NH4)2SO4...i"

"Alcohols: Simple and polyfunctional alcohols are miscible with liquid ammoina. Phenols are also soluble.

Ethers: Diethylether is moderately soluble [in liquid ammonia]. Ethers having higher molecular weights are not very soluble.

Hydrocarbons: Alkanes are insoluble, while alkenes and alkynes are slightly soluble. Benzene dissolves readily."

"All solutions of metals in liquid ammonia are metastable, though they can be stored for long periods in the absence of catalysts (impurities). Catalysts and in paticular finely divided metals (platium asbestos, platium sponge, and raney nickel), favour decompostion in accordence with: [where x is a number]

M + x NH3 --> M(NH2)x + (x/2)H2

This decompostion is used for the preparation of alkali and alkaline earth metal amides (amide reaction). It corresponds to the reaction of alkali metals with water. The catalytic activity of many metal salts (particularly iron salts) is due to the fact that the salt is first reduced and the resulting finely divided metal catalyses the amide formation."


Solublities of group 1 and 2 metals in solvents

This is (between the __ lines) generated from an article called "The Solublity of Alkali Metals in Ethers" in the Journal of the Chemical Society, April 1959, page 3767.

As the solutions were unstable to air they were handled in vacuo or in oxygen free nitrogen methods.

Dimethyl Ethylene gycol ether. Na/K - very soluble K - moderate Na, Li, Ca - None Ethyl Methyl Ethylene Glycol ether Na/K - slight Dimethyl Diethylene glycol ether Na/K - very soluble Diethyl Diethylene Glycol ether Na/K - slightly Ethyl Methyl Diethylene Glycol ether Na/K - Moderate Methyl n-propyl Diethylene Glycol ether Na/K - slightly (a) n-Butyl Methyl Diethylene Gycol ether Na/K - very slightly (a) TetraHydroFuran Na/K - slightly 1-Methoxymethyltetrahydrofuran Na/K - Very soluble 1-ethoxymethyltetrahydrofuran Na/K - moderate (b) 2-Methyltetrahydrofuran Na/K - very slightly (a) Dioxan Na/K - None Cyclic tetramer of propylene oxide Na/K - Very soluble 1: 2-Dimethoxypropane Na/K - very slightly (a) Triethylene glycol dimethyl ether Na/K - very soluble Tetraethylene glycol dimethyl ether Na/K - very soluble Ethylenediamine Na/K - very soluble Methoxyethylamine Na/K - very soluble
  1. (only on proplonged cooling to 193 K)
  2. ( slightly at room temp; moderate on cooling to 193 K)

This article goes on to state:

Lithium - doesn't dissolve in dimethylamine but is quite soluble in ethylamine

"(d) CH3-(OCH2CH2)n-OCH3 As n increases from 1 to 4 there is a very marked increase in solublity owing to an increase in the number of donor atoms per ether molecule. Entropy effects of chelation will evidently become more pronounced as n increases."


From "Ionizing Solvents", by I. Junder, yr 1970, page 39

  • Lithum dissloves 10.9 gr in 100 gr of NH3 at -33 C
  • Sodium dissolves 24.8 gr in 100 gr of NH3 "
  • Potassium dissolves 46.4 gr in 100 gr of NH3 "

U.S. Patent # 5675038 states:

Alkali metals dissolve in polyamines. For example: EDA, ethylenediamine, H2N-CH2CH2-NH2.


From "Non-aqeous Solvents", by John R. Chipperfield, copyright 1999, page 60

"HMPA (hexamethylphosphoramide) dissolves group 1 metals in a similiar way to liquid ammonia, amines and ethers, forming deep blue solutions, but only low concentrations of solvated electrons can be prepared."