Abstract
The direct conversion of solid-supported carboxylic acid allyl esters to carboxamides through the use of phenylsilane and catalytic Pd under mild reaction conditions is reported. The use of this methodology for the generation of a 48 compound solid-phase array is described herein.
1. Introduction
The synthesis of large combinatorial libraries of low molecular weight, drug-like molecules requires robust chemistry employing a wide variety of diversity elements on solid support. The carboxylic acid functional group has been widely used in solid-phase chemistry, especially when protected as an ester. An allyl ester is a commonly used protecting group that can be used with many acid or base-labile linkers, and is removed easily with Pd(PPh3)4 in various solvent systems with the aid of a scavenger reagent [1–3]. Phenylsilane, acting as a hydride donor, has been reported to be an excellent scavenger when used in conjunction with Pd(PPh3)4 in the removal of the allyl ester group [4]. Recently, it was also reported from our laboratory that phenylsilane could be directly used as an active amidation reagent of carboxylic acids like other coupling agents [5]. Based on this finding, a novel, one-step to convert allyl ester to amide, using palladium and phenylsilane as activating agents, has been developed on solid support (Scheme 1). In an effort to explore the generality and scope of this method, this reaction was examined using various structurally diverse carboxylic acids and amines (primary, secondary, or anilines) on solid support. A general method to protect carboxylic acids with an allyl ester group on solid support is also reported.
 |
2. Results and Discussion
(1) Three representative allyl ester resins (Scheme 2, 4–6) were initially prepared from the resin-bound N-benzoyl carboxylic acids 1–3, which could be easily prepared through standard procedures. Resin bound allylation proceeds well under optimized reaction conditions (5 equivalents of allylbromide, 5 equivalents of CsF, DMF, room temperature).
 |
Typical Procedure
(a) The acid resin 1–3 (2 g, ~1.0 mmol/g loading) was swollen with ~15 mL DMF at room
temperature and then 5 equiv of allybromide
and 5 equiv of CsF were added into the resin mixture. The reaction was allowed
to agitate overnight. After that, the reaction was washed with DMF(2x),
THF(3x), DCM(4x). The resin 4–6 was
dried under high vacuum pump. The resin was checked by TFA cleavage, and the
product 7–9 had 90~96% of purity and
96~100% of yield.
(b) A 48-compound array was constructed from a matrix of these three R2 allyl esters crossed with 16 R3 amines and anilines. The process was
carried out using IRORI MicroKans, which is Rf-encoded split pool synthesis
technology. The reactions were performed on a 0.02 mmol scale (0.02 mmol/one
microkan) in anhydrous CH2Cl2 at room temperature for 24
hours with 10 equivalents of amines or anilines, 20 equiv of PhSiH3 and 0.05 equiv of Pd(PPh3)4. After that, all microkans
were pooled together and then washed with DMF(2x), THF(3x), DCM(4x). All
microkans were dried under high vacuum pump and then sorted into IRORI cleavage
station. The final products 10 were
cleaved into 96 well-format plate with 30% of TFA solution in DCM. The solvent
was removed and the products were directly analyzed with flow NMR and LC-MS.
Overall yield was calculated based on the initial loading of resins.
Following
cleavage from the resin with TFA, nearly quantitative yields of the allyl ester
products 7–9 (based on the initial loading
of resin 1–3) were obtained with
high purity (>90%) (Scheme 2).
Using resins 4–6, a 48 compound array was
constructed from a matrix of these three R2 allyl esters crossed
with 16 R3R4NH amines and anilines (see Typical
Procedures above). The reactions were performed on a 0.02 mmol scale in
anhydrous CH2Cl2 at room temperature for 24 hours with 10
equivalents of amines or anilines, 20 equiv of PhSiH3 and 0.05 equiv
of Pd(PPh3)4. The final products 10 were obtained after cleavage from the resins. This entire
process was carried out using commercially available IRORI MicroKans [6] (Scheme 3).
Solid-phase
array results are provided in Table 1 below. The general reaction conditions employed worked equally well for both
primary and secondary amines with the various allyl esters. Additional experiments have indicated that
the best solvents for this method are DMF or NMP and the reaction could be
carried out with 5 equivalents of amines and 10 equiv of PhSiH3.
Direct analyses
of the cleaved products (HPLC and LC/MS) indicated high purity and yields in
most cases. All calculations were based on the initial loading of resin 1–3. All reactions were clean with the
major side product being the unprotected carboxylic acid. With an increase in
steric congestion of the amine component (entries 3, 10, 19, 26, and 35), more severe conditions were required to drive reactions to
completion (i.e., higher reaction temperatures and the use of additional
equivalents of amines). Reactions with anilines were disappointing with few
amide products obtained and high recovery of the unprotected acids observed
even under more forcing reaction conditions. No product was formed in the
reaction without addition of both PhSiH3 and catalytic Pd(Ph3P)4 reagents. It was shown in our controlling experiments, entry 38.
 |
3. Conclusion
In summary, a Pd(0)/PhSiH3 system has been successfully applied to convert resin-bond allyl esters to amides. The reaction can typically be carried out in a single step at room temperature. A systematic investigation of amine and aniline inputs has demonstrated that in general, primary amines and unhindered secondary amines give excellent yields of amides with high purity. Higher reaction temperatures and additional equivalents of amines can be used to push reactions to completion. This methodology has been recently used in a solid-phase sequence to prepare a 10000-compound library directed at the identification of protease inhibitors.