G. Rassias, S. A. Hermitage / Tetrahedron Letters 50 (2009) 5565–5568
5567
this 95:5 mixture of 18–17 was treated with 7 M NH3 in MeOH, com-
plete reaction was observed within a few hours and the desired pri-
mary alkyl amine 15 was isolated in 70% yield. This clearly
demonstrates that both requirements for rapid equilibration be-
tween 17 and 18 and fast SN2 reaction of 17 with ammonia operate
in thissystem. The stereochemicalintegrity of thestereogeniccentre
adjacent to the sulfur atom was preserved through an overall double
inversion process. The undesired isomer 16 was formed in consis-
tently small amounts regardless of the leaving group associated with
the secondary activated species 12, 14 and 18 which suggests that
some regiochemical leakage was probably occurring through 6a.
Having established that the bromide species are suitable for the
amination reaction we then focused on installing the acetamidine
group. In principle, the amidine functionality present in 5 can be
installed directly by the displacement of bromide in the equilibrat-
ing mixture 17/18 by acetamidine. This approach did not work un-
less strong bases (methoxide) were included in order to generate
the amidine free base. Such conditions were not tolerated well
by our substrates and degradation became more dominant than
amidination.11 Having established that direct amidination was
problematic, we decided to examine a stepwise approach. We ini-
tially found that amine 15 could be converted into 19a when trea-
ted with acetamidine hydrochloride in a methanolic solution of
ammonia. The minor amine isomer 16 does not appear to amidi-
nate under these conditions hence this transformation also served
as a purification step. We then demonstrated that it was possible to
achieve a tandem amination/amidination of 17 in a single-pot pro-
cess. The one-pot process required some tuning because the chlo-
ride anions present in the acetamidine salt (used in large excess)
gave rise to significant amounts of 14. The use of acetamidine
hydrobromide12 solved this problem but 19b could not be obtained
in crystalline form after aqueous work-up. We decided to perform
salt exchange as part of the work-up and succeeded in isolating
19c as a white crystalline solid. To the best of our knowledge,
acetamidinium salts have not been reported as amidination re-
agents to date. An advantage of the in situ amidination was the
suppression of 20 which probably arises from the reaction of 15
with 17. In the non-telescoped amination reaction, impurity 20
forms in approximately 1–5%, but in the presence of acetamidini-
um salts (telescoped process) the primary amine 15 amidinates
rapidly and therefore polyalkylation is avoided.
established equilibrium involving 15, 19 and 21. At low levels of
the acetamidinium salt reagent, the higher amidine 21 was formed
at significant levels. An authentic sample of 21 was prepared and
was shown to revert back to 19 and 15 when treated with 7 M
methanolic ammonia. As a result of this equilibrium, the yield of
19 was limited to 50% but work is in progress to understand this
equilibrium and implement it more effectively in the one-pot
process.
In summary, our route starts with the reduction of acid 9 to
alcohol 10, whose solution is taken forward into the activation step
resulting initially in equilibrating mesylates 11 and 12 and eventu-
ally into the equilibrating mixture of bromides 17 and 18, all pre-
sumably via thiiranium intermediate 6a. After aqueous work-up,13
the bromide species is treated with a solution of acetamidine
hydrobromide in 7 M ammonia in methanol followed by work-up
with aqueous ammonium tetrafluoroborate to furnish 19c in 40%
yield as a crystalline solid. This process spans six forward steps
(in theory operating at an average yield of 85% each) and three sets
of balanced equilibria, to deliver 19c of comparable quality to 19a
obtained by our current supply route.8 Finally, global deprotection
of 19c with phosphoric acid under carefully designed conditions,
allows isolation of our iNOS inhibitor candidate as the monophos-
phate monohydrate salt 5 in 65% yield.14
To conclude, we have successfully synthesised iNOS inhibitor 5
via a short, cost effective route which pivots on an unprecedented
counteranion-controlled double equilibrium mediated by an
unsymmetrical thiiranium species.15 We have also shown that
acetamidine salts react readily with amines to form substituted
acetamidines. We are currently looking to expand on the amidin-
ation reaction using these alternative reagents due to their inher-
ent stability, safety and low cost.
References and notes
1. For a review on episulfonium salts, see: Beaver, M. G.; Billings, S. B.; Woerpel, K.
A. Eur. J. Org. Chem. 2008, 771.
2. Al-Masoudi, N. A. L.; Hughes, N. A. J. Chem. Soc., Perkin Trans. 1 1987, 2062.
3. Bozo, E.; Boros, S.; Kuszmann, J.; Gacs-Baitz, E. Carbohydr. Res. 1996, 290, 159.
4. Converso, A.; Saadi, P.-L.; Sharpless, K. B.; Finn, M. G. J. Org. Chem. 2004, 69,
7336.
5. Alderton, W. K.; Angell, A. D. R.; Craig, C.; Dawson, J.; Garvey, E.; Moncada, S.;
Monkhouse, J.; Rees, D.; Russell, L. J.; Russell, R. J.; Schwartz, S.; Waslidge, N.;
Knowles, R. G. Br. J. Pharmacol. 2005, 145, 301.
6. Use of commercial borane in THF offered a good reaction and easier work-up
but the product was contaminated with varying amounts of n-butanol
(resulting from the reductive ring-opening of THF), depending on the date of
manufacture and storage conditions of the reagent.
7. Alternative chlorinating agents such as PPh3/NCS, SOCl2 and PPh3Cl2 gave
similar results with respect to the distribution of the chloride species and
showed the same insensitivity towards base, temperature and solvent.
8. For the current supply route, see (a) Rassias, G.; Hermitage, S. A.; Sanganee, M.
J.; Kincey, P. M.; Smith, N. M.; Andrews, I. P.; Borrett, G. T.; Slater, G. R. Org. Proc.
Res. Dev. 2009, 13, 774–780; (b) Hermitage, S. A.; Panchal, T.; A.; Rassias, G.;
Sanganee, M. J. PCT Int. Appl. 2005, WO 2005005377; Chem. Abstr. 2005, 142,
114456.
9. Use of azide chemistry was deliberately avoided due to scale-up and safety
concerns and because of the poor performance of these thioether derivatives in
the hydrogenation reactions we attempted previously (see Ref. 8). HMDS,
hexamethylene triamine and ammonium salts in conjunction with bases failed
to act as ammonia equivalents in this reaction. 7 M Ammonia in methanol
outperformed the more dilute 2 M alternative and solutions of ammonia in
water, ethanol, isopropanol and dioxane.
10. Attempts to prepare the mesylate species under chloride-free conditions using
methanesulfonic anhydride also failed to improve the impurity profile. In the
absence of other nucleophilic species the base required for this reaction
invariably engaged the mesylate(s) and quaternary ammonium salts started to
predominate for all bases used (Et3N, DIPEA, Py). Quaternary salt formation
was greatly limited by the use of 2,6-lutidine, but disappointingly, the parent
bases could not be displaced from their quaternary derivatives by ammonia.
11. For the use of bis-Cbz protected acetamidine as nucleophile, see: Eustache, J.;
Grob, A. Tetrahedron Lett. 1995, 36, 2045; Eustache, J.; Grob, A.; Lam, C.; Sellier, O.;
Schulz, G. Bioorg. Med. Chem. Lett. 1998, 8, 2961. This reagent would generate the
Cbz-protected amidine. Hydrogenolysis of the Cbz-protected 15 was one of the
weaknesses identified in our first route due to the poisoning effect of the sulfur
NHBoc
O
NH.HX
a X=Cl
b X=Br
c X=BF4
OtBu
S
19
N
H
NHBoc
NHBoc
tBuO
OtBu
S
S
S
N
H
O
O
O
20
NHBoc
NHBoc
O
tBuO
OtBu
S
N
N
H
21
During our attempts to optimise the new amidination reaction
independently using authentic samples of 15 and acetamidine
hydrochloride, we found that this process was actually a rapidly
Rassias, Geracimos
