ECHET96 Search CD [Molecules: 10] [Related articles/posters: 052 073 018 007 060 ]

email.gif - 0.3 KECHET96 Article 019 P.S.Andrada
email.gif - 0.3 KRe: ECHET96 Article 019 Mario R. Tellez

Synthesis of 2-(2-fluorophenylimino-3-(2-fluorophenyl)-
1,3-thiazolidine-4,5-dione, di-4-fluorophenyl-thioparabanic acid, and 3-(p-fluorophenyl)-1,3-thiazetidin-4-one-2-
(p-fluorophenyl)-imine. Easy and versatile conversion of the thiazetidinone into carbodiimide or isourea derivatives

Mario R.Tellez

New Mexico State University, Jornada Experimental Station, USDA-ARS box 3JER, Las Cruces, NM 88003 USA

In the course of
our studies on nitrenes generated from N-arylsulfonyloxy-N'-4-fluorophenyl-O-alkylisoureas we became interested in the synthesis of N,N'-bis(4-fluorophenyl)-O-methyl isourea 1. We hereby report on the synthesis through known methods<66AG(E)704> of 2-(2-fluorophenylimino-3-(2-fluorophenyl)-1,3-thiazolidine-4,5-dione 2, di-4-fluorophenylthioparabanic acid 3 and 3-(p-fluorophenyl)-1,3-thiazetidin-4-one-2-(p-fluorophenyl)-imine 4 and the utility of this later compound in the synthesis of both carbodiimides and isoureas.

Treatment of thioureas with either phosgene (Scheme 1) or oxalyl chloride is known to give amido chlorides in good yields<60AG(E)48>, <66AG(E)704>.

Scheme 1

Synthesis of this amido chloride would then open two routes to the desired isourea. A carbodiimide could be obtained initially by hydrolysis of N,N'-disubstituted carbamido chlorides as observed by Eilingsfeld, Seefelder, and Weidinger<60AG(E)48> (Scheme 2) and then be converted to the isourea by treatment with an appropriate alcohol and catalyst<67CRV107>, <81CRV589>


Scheme 2

Eilingsfeld and coworkers<64CB1232> also observed that reaction of the amido chlorides with two equivalents of alkoxide gives the corresponding urea acetals (Scheme 3).

Scheme 3

It was thought that by proper control of the reaction conditions the reaction could be stopped at the isourea stage (Scheme 4), isoureas being the key intermediates towards the synthesis of the nitrene precursors<95MI54-ORGN>.

Scheme 4

Treatment of the thiourea 5<53LA77>, <55JCS1573> with oxalyl chloride (Scheme 5) afforded a mixture of 2-(2-fluorophenylimino-3-(2-fluorophenyl)-1,3-thiazolidine-4,5-dione 2 and di-4-fluorophenylthioparabanic acid 3 instead of the expected amido chloride<60AG(E)48>. This result is in accordance with findings by Ulrich and Sayig<66AG(E)704> indicating that thioureas give the corresponding thiazolidinediones when treated with oxalyl chloride. They also observed that these thiazolidinediones are converted into the parabanic acid derivative when warmed in ethanol.

Scheme 5

Heating also converted our thiazolidinedione 2 to 3. We were able, however, to convert our mixture completely into the thioparabanic acid 3 derivative without heating by treatment first with 25% sodium hydroxide and then with aqueous hydrochloric acid.

Attempts to convert the thiourea 5 to the corresponding amidochloride with phosgene<64CB1232> were not successful either. Ulrich and Sayig<66AG(E)704> observed that N,N'-aryldisubstituted thioureas, when treated with phosgene, give the amidochloride along with some thiazetidinone as a byproduct. In our case, instead, 3-(p-fluorophenyl)-1,3-thiazetidin-4-one-2-(p-fluorophenyl)-imine 4 (Scheme 6) was obtained in 63% yield as the main product from the reaction of N,N'-bis(4-fluorophenyl)thiourea 5 with one third of an equivalent of triphosgene [bis(trichloro-methyl)carbonate] (one third of an equivalent was used since triphosgene generates three equivalents of phosgene).

Scheme 6

The GC/MS of the thiazetidinone 4 proves to be interesting. It affords three sharp GC peaks which correspond to scission of the four-membered ring along both axis to afford on the one hand the carbodiimide 6 (and pressumably OCS), and on the other hand, the isothiocyanate 7 and isocyanate 8 (Scheme 7):

Scheme 7

Heating of a thiazetidinone is reported in the literature<66AG(E)704> to afford the corresponding carbodiimide. Heating the thiazetidinone 4 either neat or in refluxing toluene (Scheme 8) affords, as expected, the carbodiimide 6<53LA77> as the sole product (presumably with evolution of O=C=S).

Scheme 8

A mixture of the starting material 4 and carbodiimide 6 was obtained when 4 was heated to reflux in dichloromethane containing one and a half equivalents of ethanol.

N,N'-bis(4-fluorophenyl)-O-ethylisourea 1 could be obtained, in the absence of any catalyst, by refluxing the carbodiimide 6 in absolute ethanol (Scheme 9) overnight.

Scheme 9

Synthesis of carbodiimide 6 was also performed through the more classical desulfurization method using lead oxide<53LA77> in refluxing acetone. This reaction, however, failed inexplicably on several occasions giving traces of the corresponding N,N'-bis(4-fluorophenyl)urea 9 and large quantities of a clear colourless oil and with no sign of the desired carbodiimide.

On the other hand, the N,N'-bis(4-fluorophenyl)-O-ethylisourea 1 could be obtained in good yield directly from the thiazetidinone by refluxing 3-(p-fluorophenyl)-1,3-thiazetidin-4-one-2-(p-fluorophenyl)-imine 4 in ethanol (Scheme 10) overnight.

Scheme 10

This isourea 1 is a rather stable oil which remains mostly intact even on prolonged storage, showing only a minimal conversion to the urea (~0.3%) after one year.

Although neither the use of triphosgene nor oxalyl chloride afforded the desired amido chlorides, we have found the thiazetidinone 4 to be a highly versatile intermediate allowing for the facile conversion of thioureas to either carbodiimides or isoureas in good yields. An alternative method for the conversion of thiazolidinediones into their corresponding thioparabanic acid rearrangement products is also presented.


Back to top of poster