Triphenylphosphonium Anhydride Trifluoromethanesulfonate

[72450-51-2]  · C38H30F6O7P2S2  · Triphenylphosphonium Anhydride Trifluoromethanesulfonate  · (MW 838.76)

(oxophilic reagent for net removal of water in various reactions1,2)

Physical Data: highly hygroscopic, white crystalline solid.3

Solubility: slightly sol chlorocarbon solvents; insol Et2O; reacts with THF.

Preparative Methods: triphenylphosphine oxide in CH2Cl2 is treated with Trifluoromethanesulfonic Anhydride in CH2Cl2 at 0 °C. A precipitate of the reagent forms in <15 min.

Handling, Storage, and Precautions: prepared directly before use.


The reagent is a potent oxophile, forming triphenylphosphonium ethers (Ph3+P-O-R) with OH groups and Triethylamine (or preformed oxide anions). These activated ethers may be either displaced or eliminated by nucleophiles or bases, respectively, with triphenylphosphine oxide as the leaving group. The hydroxyl substrate is added to a slurry of the reagent and rapid formation of the ether is indicated by solution of the reagent. Subsequent reactions are usually fast, although higher temperatures are available using ethylene dichloride as solvent. Passage of the reaction mixture through a short silica column usually readily entraps the phosphine oxide. The reagent is generally unreactive to carbonyl groups or ethers (except epoxides, below) and is quite selective for attack at oxygen over nitrogen nucleophiles.

Carboxylic Acids.2

Usually, dehydration of acids to anhydrides is quickly achieved by adding a solution of the acid and TEA to the reagent slurry (2:2:1 = acid:TEA:reagent). When the a-CH of the acid is modestly acidic, however, dehydration proceeds to the ketene, as in a simple preparation of diphenylketene from diphenylacetic acid. In either case the net result is an overall dehydration of the substrate(s).

Ester formation with primary alcohols is similarly facile, requiring 15 min at room temperature with the mixed acid and alcohol, and is equally effective in acylating phenols. Since the phenol may be activated first with the reagent, its ready O-acylation implies a rapid equilibration of the activated species, as in eq 1. With t-butyl alcohol, however, the yield is only 42%, implying a large extent of elimination.

The same equilibration appears in amide formation on adding mixed acid and primary or secondary amine to the reagent. While Ph3+P-NHR may form, it exchanges to activate the acid and then yields the amide.2 The preference for Ph3+P-O-R over Ph3+P-NR2 in the exchange may arise from the presence of two electron pairs on oxygen to back donate into the phosphorus d-orbitals, instead of only one on nitrogen. In a similar vein, amides and amines yield amidines on treatment with the reagent (eq 2; POP = reagent), although the reaction requires some 8 h at room temperature. The reagent has been used in the same fashion to convert sulfonic acid salts to their esters and amides in the synthesis of thymidine dimer analogs with sulfonic linkages.4

The two operations may occur successively with diamines, as in the double dehydration of benzoic acids with phenylene diamines to form benzimidazoles (eq 3; Z = NH). The conversion is complete in less than an hour at rt and appears much superior to traditional methods with higher temperatures and competing diamide formation. Similarly, benzoxazoles are conveniently formed from o-aminophenols and benzothiazoles from o-aminothiophenols (eq 3; Z = O and S, respectively)2. A six-membered perimidine was formed from 1,8-diaminonaphthalene and benzoic acid.

Internal cyclization of phenylalkanoic acids occurs with the reagent without bases present; the expelled triflic acid acts as a strong acid for activation, as in eq 4, to afford indanones, tetralones, and benzosuberones.2 A case of acid-catalyzed Claisen condensation was also encountered with phenylacetic acid (eq 5).

Other Substitutions.

With the exception of ketene formation, all the reactions above are substitutions at carboxylic carbons. With simple alcohols, activation has the advantage of affording the phosphonium ether without any nucleophile present except the inactive triflate; hence other nucleophiles of choice may be added for displacement. With primary and acyclic secondary alcohols, reduction with Sodium Borohydride (as a slurry in CH2Cl2) occurs at rt in a few hours to give the corresponding alkane.5 Although Mitsunobu conditions presumably generate the same phosphonium ethers, borohydride reduction did not occur. On activation of cyclohexanols with the reagent, however, borohydride only generates Ph3P.BH3 and the alcohol is unchanged, apparently owing to steric hindrance. Although azide ion displaces the activated alcohols, halide ions are too unreactive.6

The internal displacement of the phosphonium ether is seen in the reaction of diols or amino alcohols with the reagent to form cyclic ethers or amines, respectively (eq 6), with three-, five-, or six-membered rings (the 1,3-diols for four-membered rings fragment to alkene and aldehyde).7 The amino alcohols again imply the facile N/O exchange of the phosphonium group.


Without nucleophiles present, elimination occurs instead, as with diphenylacetic acid to ketene above. Unsubstituted amides and aldoximes give nitriles at rt in minutes.8 Tertiary alcohols eliminate with ease and phosphonium ethers of secondary alcohols eliminate with K2CO3 or TEA on warming in CH2Cl2. Regiochemistry is not yet clear but there is an implication of favored cis elimination, for menthol gives both D2- and D3-menthenes8 (cis required for the former) while only cis elimination occurred in two tropinol isomers9 and a decalin case.10

Two other eliminations, however, have new synthetic value. Ketones activated by b-carbonyl or aryl groups are easily enolized and the reagent then serves to dehydrate the ketone to an alkyne in good yield (eq 7).11 Dehydration of simple ketones to alkynes has not yet been perfected.

Although epoxides are easily formed from 1,2-diols with one mole of reagent, the epoxide can react further with another mole, dehydrating to the dienes.12 This appears to provide a very convenient path for converting alkenes to dienes by successive epoxidation and dehydration, as in the case of limonene (eq 8).


The groups on phosphorus need not be phenyl; the related tributyl derivatives have been similarly used13 to make 1,2-cis-furanoside ethers14 and also as electrophilic catalysts for aldol and Michael reactions of silyl enol ethers.15 To avoid the chromatographic separation of the product triphenylphosphine oxide, a basic variant of the reagent was created in the same way from Ph2PO-N(CH2CH2)2NMe, which is then returned from the reaction as its salt and easily extracted into water.2

The triphenylphosphonium ethers are also regarded as intermediates in the Mitsunobu reaction, prepared from Ph3P + (RO2C-N=)2 to activate an alcohol for substitution (see Triphenylphosphine-Diethyl Azodicarboxylate).16 They were earlier prepared electrochemically from Ph3P + ROH and used for alkylation of imidazole and thiophenol17 and carboxylic acids,18 or applied to amides and ureas for dehydration to nitriles and carbodiimides, respectively, with bases.19 The acyloxytriphenylphosphonium ions were similarly prepared from Ph3P + RCO2H and converted to esters and amides as above, although in lower yields.20

1. Hendrickson, J. B.; Hussoin, Md. S. JOC 1987, 52, 4137.
2. Hendrickson, J. B.; Hussoin, Md. S. JOC 1989, 54, 1144.
3. X-ray structure: Aaberg, A.; Gramstad, T.; Husebye, S. TL 1979, 2263.
4. Reynolds, R. C.; Crooks, P. A.; Maddry, J. A.; Akhtar, M. S.; Montgomery, J. A.; Secrist, J. A. JOC 1992, 57, 2983.
5. Hendrickson, J. B.; Singer, M.; Hussoin, Md. S. JOC 1993, 58, 6913.
6. Ramos, S.; Rosen, W. TL 1981, 35; JOC 1981, 46, 3530.
7. Hendrickson, J. B.; Hussoin, Md. S. SL 1990, 423.
8. Hendrickson, J. B.; Schwartzman, S. M. TL 1975, 277.
9. Majewski, M.; Zheng, G. Z. CJC 1992, 70, 2618.
10. Blazejewski, J. C.; Leguyader, F.; Wakselman, C. JCS(P1) 1991, 1121.
11. Hendrickson, J. B.; Hussoin, Md. S. S 1989, 217.
12. Hussoin, Md. S. PhD Thesis, Brandeis, 1990.
13. Kobayashi, S.; Mukaiyama, T. J. Synth. Org. Chem. Jpn. 1990, 48, 1030.
14. Mukaiyama, T.; Suda, S. CL 1990, 1143.
15. Mukaiyama, T.; Matsui, S.; Kashiwagi, K. CL 1989, 993.
16. Mitsunobu, O. S 1981, 1; Hughes, D. L.; Reamer, R. A.; Bergan, J. J.; Grabowski, E. J. JACS 1988, 110, 6487.
17. Ohmori, H.; Nakai, S.; Sekiguchi, M.; Masui, M. CPB 1980, 28, 910.
18. Ohmori, H.; Nakai, S.; Miyasaka, H.; Masui, M. CPB 1982, 30, 4192.
19. Ohmori, H.; Sakai, K.; Nakai, N.; Mizuki, Y.; Masui, M. CPB 1985, 33, 373.
20. Ohmori, H.; Maeda, H.; Kikuoka, M.; Maki, T.; Masui, M. T 1991, 47, 767.

James B. Hendrickson

Brandeis University, Waltham, MA, USA

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