Trimethylamine N-Oxide


[62637-93-8]  · C3H9NO  · Trimethylamine N-Oxide  · (MW 75.13)

(reagent for conversion of alkyl halides and sulfonates to aldehydes and ketones; precursor for azomethine ylide useful for synthesis of N-methylpyrrolidines; reagent for decomplexation of tricarbonyldieneiron complexes)

Alternate Name: TMANO.

Physical Data: mp 213-214 °C (dec).1 Various melting points have been reported, ranging from 208 °C2 to 225-227 °C. The dihydrate, Me3NO.2H2O, is a white solid, mp 96 °C.

Solubility: sol H2O, EtOH; insol Et2O benzene, and other hydrocarbon solvents; sparingly sol hot CHCl3.

Form Supplied in: the dihydrate is commercially available as a white solid.

Preparative Methods: aqueous Trimethylamine (100 mL, 33%) is mixed with a pure 3% aqueous solution of Hydrogen Peroxide (600 mL) and is set aside at rt for 24 h. The solution is evaporated to dryness, and the product is recrystallized from EtOH-Et2O to give the dihydrate as needles, mp 96 °C.3,4 Several procedures are available for preparing the anhydrous reagent. The dihydrate is heated at 105 °C and 15 mmHg for several hours, after which it is sublimed at 180-200 °C and 15 mmHg. A more convenient procedure is to remove water by azeotropic distillation with benzene or toluene (Dean-Stark trap), or by co-distillation from a suspension of the reagent in DMF, followed by solvent removal in vacuo.

Handling, Storage, and Precautions: no reported hazards for this reagent. The anhydrous material is hygroscopic and should be prepared and used fresh or stored at rt in a vacuum desiccator over P2O5. Trimethylamine N-oxide is an oxidizer and should be stored away from reducing agents.

Preparation of Aldehydes and Ketones.

Alkyl bromides, iodides, and tosylates react with Me3NO in CHCl3 to give reasonable yields of aldehydes and ketones (eqs 1-3).5

Oxidation of Organoboranes.

Trimethylamine N-oxide is a useful alternative to alkaline Hydrogen Peroxide for the conversion of intermediates from the hydroboration of alkenes and alkynes to alcohols, especially in those cases where the product is sensitive to either high pH or H2O2. For example, cyclopropanols are reported to be unstable to alkaline H2O2, making hydroboration of cyclopropenes problematic. Oxidation of the intermediate boranes with Me3NO in hot toluene, followed by methanolysis of the derived tricyclopropylborate, gives cyclopropanols in good overall yield (eq 4).6

Later studies by Kabalka and Hedgecock7 have shown that commercially available Me3NO dihydrate is an effective reagent for organoborane oxidation, and tolerates a wide variety of functional groups. Owing to the fact that the amine oxide is not appreciably soluble in organic solvents such as THF, toluene, and CHCl3, it is recommended that the oxidation is carried out with vigorous stirring.8 Rates of oxidation are sensitive to the nature of the alkyl group (secondary cycloalkyl > secondary alkyl > n-alkyl > b-branched primary alkyl), and the second and third alkyl groups are removed more slowly. Choice of solvent does not appear to affect the rate of oxidation; the recommended reaction conditions are diglyme at reflux. Two simple examples are given in eqs 5 and 6.

A useful example of the application of Me3NO for borane oxidation is found in the formal total synthesis of (±)-perhydrohistrionicotoxin by Carruthers and Cumming (eq 7).9 Oxidation of the borane intermediate from the azaspirocycle by using alkaline hydrogen peroxide proved extremely difficult and gave a low yield of the desired alcohol (34% as a mixture of stereoisomers), this being attributed to the formation of a stable aminoborane.10 Carruthers overcame this problem by using Me3NO in refluxing diglyme, to give a mixture of stereoisomers that could be separated by chromatography of their acetate derivatives. The alcohol shown in eq 7 is a known intermediate in a synthesis of perhydrohistrionicotoxin.11

Trimethylamine N-oxide is also useful for oxidation of vinylborane intermediates formed during the hydroboration of alkynes.12 The use of hydrogen peroxide in such reactions leads to overoxidation (eq 8), and this is avoided by using Me3NO in refluxing THF in a one-pot procedure to generate acylsilanes from silylalkynes (Scheme 1).

Extension of this chemistry to the oxidation of organoaluminum derivatives has been described by Kabalka and Newton,13 which is an extension of a reaction first reported by Köster.14 It should be noted, however, that vinylalanes are not oxidized by this procedure; R1CH=CHAlR22 gives only the products of oxidation of the Al-R2 bonds, with no aldehydes being detected. Examples of the procedure are shown in eq 9.

Azomethine Ylides from Me3NO: Synthesis of N-Methylpyrrolidines.

Treatment of anhydrous Me3NO with Lithium Diisopropylamide (LDA) gives an azomethine ylide which reacts in situ with alkenes in a dipolar cycloaddition to give N-methylpyrrolidines (eq 10).15 Previous methods for generating azomethine ylides required electron-withdrawing or conjugatively stabilizing substituents.16 Reaction of the azomethine ylide with dihydronaphthalenes has been described,17 but the poor solubility of Me3NO in THF was a problem for ylide generation, and it was found that a four-fold excess of the reagent was necessary to ensure complete consumption of all the alkene (eq 11).

Decomplexation of Diene-Fe(CO)3 Complexes.

Me3NO has been used extensively in organometallic chemistry as a reagent for the selective removal of carbonyl ligands. An extension of this reactivity leads to applications in the decomplexation of diene-iron tricarbonyl complexes, which are being used increasingly as intermediates for organic synthesis. The first report of Me3NO being used for this type of ligand disengagement was by Shvo and Hazum,18 summarized in eqs 12-15. The last two examples are particularly noteworthy, since the complexes are quite sensitive to the acidic conditions that are generated when transition metal oxidants such as Iron(III) Chloride or Cerium(IV) Ammonium Nitrate are used for decomplexation. Until the development of this method the removal of Fe(CO)3 from diene complexes was considered a major obstacle to their synthetic application. The reaction conditions vary according to the diene, and must be determined empirically.

Evidence that the mechanism of the reaction involves two amine oxide molecules per diene-Fe(CO)3 comes from the isolation of an intermediate that has a carbonyl ligand replaced by Me2NH (eq 16).19a A plausible mechanistic explanation is shown in Scheme 2, though this is somewhat different from the mechanism proposed for reaction of Me3NO with Fe(CO)5.19b

A large number of examples of the use of Me3NO for decomplexation can now be found in the synthetic literature. A comparison with iron(III) chloride is revealed by the decomplexation shown in eq 17.20 While the methyl-substituted complex (X = Me) could be demetalated using FeCl3 or Me3NO, the use of FeCl3 is problematic when X = OR, SPh, or NHR, owing to the propensity of these complexes to undergo acid-promoted scission of the C-X bond to regenerate the dienyl-Fe(CO)3 cation. Me3NO is non-acidic and does not suffer this limitation.

Similar compatibility with heteroatom substituents is observed during the synthesis of (-)-gabaculine reported by Birch et al.,21 who also found that N,N-dimethylacetamide is an excellent solvent for decomplexation (Scheme 3), by Pearson and Ham10 in their formal synthesis of perhydrohistrionicotoxin (eq 18), and by Pearson and Srinivasan in some approaches to cycloheptadiene triols (eqs 19 and 20).22 The other useful aspect of Me3NO for these decomplexations is that vinyl ether groups are not hydrolyzed to ketones under the reaction conditions.

Owing to the fact that the decomplexation reaction proceeds via nucleophilic attack on a carbonyl ligand, Me3NO is not very useful for converting diene-Fe(CO)2PR3 complexes to dienes, since the poorer p-acceptor capacity of the phosphine ligand places more electron density at the carbonyls. For this reason, it has been found better to use Dipyridine Chromium(VI) Oxide in CH2Cl2 for decomplexation of the phosphine-substituted derivatives (eq 21).23

Shvo and Hazum have also reported the use of Me3NO as promoter during the conversion of dienes to diene-Fe(CO)3 complexes.24 Reaction of Fe(CO)5 with Me3NO at controlled temperature leads to the formation of a coordinatively unsaturated Fe(CO)4 derivative by a mechanism related to that shown in Scheme 2. In the presence of a diene, rapid coordination of alkene occurs, followed by a second CO loss and alkene coordination to give the diene complex (eqs 22 and 23). It should be noted that the normal conditions for Fe(CO)5 -> diene-Fe(CO)3 conversion are refluxing in di-n-butyl ether (ca. 140 °C).

This observation, coupled with the fact that diene-Fe(CO)2PR3 complexes are quite resistant to Me3NO, leads to a mild method for the conversion of diene-Fe(CO)3 to diene-Fe(CO)2PR3 complexes (eqs 24-26).22,25,26 The examples in eq 24 (R1 or R2 = OMe) are especially interesting because other methods for ligand exchange (thermal or photochemical) with these electron-rich diene complexes either fail completely or give very low yields.

Related Reagents.

N-Benzyl-N-(methoxymethyl)-N-trimethylsilylmethylamine; Cerium(IV) Ammonium Nitrate; Hydrogen Peroxide; Iron(III) Chloride; N-[(Tri-n-butylstannyl)methyl]benzaldimine.

1. Monagle, J. J. JOC 1962, 27, 3851.
2. Meisenheimer, J. LA 1913, 397, 273.
3. Weygand, C.; Hilgetag, G. Organische-chemische Experimentierkunst; English: Preparative Organic Chemistry; Hilgetag, G.; Martini, A.; Eds.; Wiley: Chichester, 1972; p 574. Hickinbottom, W. J. Reactions of Organic Compounds; Longmans: New York, 1936; p 277.
4. Anhydrous reagent: Soderquist, J. A.; Anderson, C. L. TL 1986, 27, 3961. Franzen, V. OS 1967, 47, 96.
5. Franzen, V.; Otto, S. CB 1961, 94, 1360.
6. Köster, R.; Arora, S.; Binger, P. AG(E) 1969, 8, 205.
7. Kabalka, G. W.; Hedgecock, H. C., Jr. JOC 1975, 40, 1776.
8. Kabalka, G. W.; Slayden, S. W. JOM 1977, 125, 273.
9. Carruthers, W.; Cumming, S. A. CC 1983, 360.
10. Pearson, A. J.; Ham, P. JCS(P1) 1983, 1421.
11. Corey, E. J.; Balanson, R. D. H 1976, 5, 445, and references cited therein.
12. Miller, J. A.; Zweifel, G. S 1981, 288.
13. Kabalka, G. W.; Newton, R. J., Jr. JOM 1978, 156, 65.
14. Köster, R.; Morita, Y. LA 1967, 704, 70.
15. Beugelmans, R.; Negron, G.; Roussi, G. CC 1983, 31.
16. (a) Huisgen, R. AG(E) 1980, 19, 947. (b) Kellogg, R. M. T 1976, 32, 2165.
17. De, B.; DeBernardis, J. F.; Prasad, R. SC 1988, 18, 481.
18. Shvo, Y.; Hazum, E. CC 1974, 336.
19. (a) Eekhof, J. H.; Hogeveen, H.; Kellogg, R. M. CC 1976, 657. (b) Elzinga, J.; Hogeveen, H. CC 1977, 705.
20. Shu, B. Y.; Biehl, E. R.; Reeves, P. C. SC 1978, 8, 523.
21. Bandara, B. M. R.; Birch, A. J.; Kelly, L. F. JOC 1984, 49, 2496.
22. Pearson, A. J.; Srinivasan, K. JOC 1992, 57, 3965.
23. Pearson, A. J.; Lai, Y. S.; Lu, W.; Pinkerton, A. A. JOC 1989, 54, 3882.
24. Shvo, Y.; Hazum, E. CC 1975, 829.
25. Birch, A. J.; Kelly, L. F. JOM 1985, 286, C5.
26. Howell, J. A. S.; Squibb, A. D.; Goldschmidt, Z.; Gottlieb, H. E.; Almadhoun, A.; Goldberg, I. OM 1990, 9, 80.

Anthony J. Pearson

Case Western Reserve University, Cleveland, OH, USA

Copyright 1995-2000 by John Wiley & Sons, Ltd. All rights reserved.