amount of the thiazoline cyclization product was observed
(Scheme 1). To confirm that thio-CDI was indeed promot-
With the initial proof of concept in hand, we looked at
optimizing the reaction with the addition of an auxiliary base.
Bases such as KOtBu, Cs2CO3, K2CO3, Et3N, imidazole, and
DBU were screened, all at 2 equiv in relation to the thiourea
starting material (5). We found that, in general, the inorganic
bases were detrimental to the reaction (low assay yields with
the balance of the material being unidentified impurities as
well as residual starting material), whereas the weaker amine
bases had little to no effect on the yield. The performance
of the reaction was also assessed using solvents other than
THF. Acetone, toluene, CH2Cl2, and EtOAc perform com-
parably to THF, while more polar solvents (DMF, NMP,
and acetonitrile) gave lower yields. On the basis of these
results, it was determined that our preferred conditions for
this transformation are to use THF at ambient temperature
with no additional base.
Scheme 1
.
Discovery of Thiazoline Formation under Novel
Conditions
Although the reaction conditions described above were
successful in converting hydroxyethyl-thioureas to the cor-
responding thiazolines under mild conditions, the use of thio-
CDI on a larger scale can be problematic. Thio-CDI is fairly
expensive compared to other activating reagents and has
limited availability in large batch sizes. It is a foul-smelling
solid that is unstable to moisture, which leads to storage
problems. For these reasons, we decided to investigate
alternative reagents to promote the cyclization.
Several common activating agents were screened using
compound 5 as a test substrate to evaluate their reactivity
and selectivity compared to the results using thio-CDI.
Several cyclization products (6-9) as well as a substitution
product (10) were observed as detailed in Table 1.9 The
use of diethylazodicarboxylate (Mitsunobu conditions,
entry 1) gave a mixture of S and N cyclization products,
with imidazolidinethione 9 being favored 3:1 over thia-
zoline 6. Activation of the thiourea using MeI or tosyl
chloride exclusively gave cyclization to oxazoline 7, and
although mesyl chloride was also selective for oxazoline
7, 10% of thiazoline 6 was observed (entries 2-4). The
results in entries 1-4 all corroborate the chemoselectivi-
ties reported in the literature for similar substrates.6,7,10
Other activating agents that are novel to this type of
transformation were also examined with mixed results. Both
triphosgene and thiophosgene produced multiple cyclization
products; however, the major product with these reagents
was alkyl chloride 10.
ing cyclization, thiourea 1 was purified and resubjected
to the reaction conditions with a full equivalent of thio-
CDI. The reaction proceeded readily at room temperature,
and phenylaminothiazoline 2 was isolated in 54% yield.
This serendipitous result led us to further explore the scope
and general utility of this novel transformation.
We chose N-phenylthioureas with simple alkyl-substi-
tuted amino alcohol side chains to study this transforma-
tion. These substrates can be easily prepared by reaction
of phenylisothiocyanate (3) with the appropriate amino
alcohol in THF at ambient temperature.8 The model
substrate used for optimization of the methodology
(thiourea 5, Scheme 2) was prepared in 93% yield from
Scheme 2. Synthesis of a Model System Using Thio-CDI
Performing the reaction using Vilsmeier reagent gave
60% of thiazoline 6 with the remaining 40% being comprised
of multiple unidentified impurities. As expected, thio-CDI
gave thiazoline 6 as the major product with only 3% of
oxazoline 7, but we were pleased to see that CDI gave
exclusively thiazoline 6. As the reaction profile for CDI is
equal to or superior to thio-CDI, and CDI is not hampered
by the limitations previously listed for thio-CDI, this reagent
was chosen for further examination.
phenylisothiocyanate (3) and valinol (4). To confirm the
result observed with compound 1, thiourea 5 was treated
with thio-CDI which gave the desired product, thiazoline
6, in 61% isolated yield.
(9) The products were characterized using 1H NMR and LC-MS
analysis, and the product ratios were generated from the 1H NMR
integrations of the crude reaction mixtures.
(8) (a) Kruse, L. I.; Kaiser, C.; DeWolf, W. E., Jr.; Frazee, J. S.; Garvey,
E.; Hilbert, E. L.; Faulkner, W. A.; Flaim, K. E.; Sawyer, J. L.; Berkowitz,
B. A. J. Med. Chem. 1986, 29, 2465–2472. (b) Lattanzi, A. Synlett 2007,
13, 2106–2110.
(10) (a) Perez, E. M.; Oliva, A. I.; Hernandez, J. V.; Simon, L.; Moran,
J. R.; Sanz, F. Tetrahedron Lett. 2001, 42, 5853–5856. (b) Hetenyi, A.;
Szakonyi, Z.; Klika, K.; Pihlaja, K.; Fulop, F. J. Org. Chem. 2003, 68,
2175
.
Org. Lett., Vol. 12, No. 23, 2010
5527