Platinum-catalyzed amidocarbonylation

Article abstract of DOI:10.1246/cl.2003.160

The first example of platinum-catalyzed amidocarbonylation of aldehydes with amides and carbon monoxide is described. In contrast to precedent palladium catalysis, a remarkable ligand acceleration by phosphines was observed. Furthermore, an optically active N-acetyl amino acid was partially epimerized under the platinum-catalyzed conditions, while faster racemization was observed under the palladium catalysis.

Full text of DOI:10.1246/cl.2003.160

160
Chemistry Letters Vol.32, No.2 (2003)
Platinum-catalyzed Amidocarbonylation
Takahiro Sagae, Masaharu Sugiura, Hiroyuki Hagio, and Shu¯ Kobayashiꢀ
Graduate School of Pharmaceutical Sciences, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033
(Received September 17, 2002; CL-020802)
The first example of platinum-catalyzed amidocarbonylation
the similar conditions. With PtCl2(cod) (5 mol%), sterically
hindered monophosphines (10mol%), 2-(dicyclohexylphosphi-
no)biphenyl (biphPCy2),9 PCy3, and Pt-Bu3 were less effective
than biphPt-Bu2 (3: 22%, 7%, and 13%, respectively). However,
simple PPh3 afforded 3 in good yield (45%). On the other hand,
bidentate ligands, DPPE and BINAP suppressed the reaction
completely. With PtX2(cod)/biphPt-Bu2 (X = Br or I), the
catalytic activity was decreased (3: 59% and 12%, respectively).
This suggests that the halide anion coordinated to the metal was
not replaced by the added LiBr. The combination of PtCl2 with
biphPt-Bu2 showed comparable activity to that with PtCl2(cod)
(3: 58–75%). Therefore, the effect of triarylphosphines was
extensively investigated using PtCl2. It was found that (1)
biaryldiphenylphosphines, 2-(diphenylphosphino)biphenyl and
2-(diphenylphosphino)-20-methoxy-1,10-binaphthyl (MOP),10
were ineffective (3: 0%), (2) para-substituted triarylphosphines,
P(p-MeC6H4)3 and P(p-ClC6H4)3, were effective regardless of
their electronic nature (3: 45% and 54%, respectively), (3) mono-
ortho-substitution or di-metha and para-substitutions on tri-
arylphosphines, P(o-MeC6H4)3, P[3,5-Me2-4-(MeO)C6H3]3, and
P(3,5-Me2C6H3)3, were effective (3: 62%, 43%, and 61%,
respectively), and (4) bis-ortho-substitution, P(2,4,6-Me3C6H2)3
and P[2,6-(MeO)2C6H3]3, decreased the catalytic activity sig-
nificantly (3: 8% and 0%, respectively). Consequently, it was
concluded that the platinum catalyst system tended to depend on
steric factors of the triarylphosphines rather than electronic ones.
Further investigations of the platinum catalysts revealed that
PtCl2/biphPt-Bu2 system promoted the reaction even without the
addition of H2SO4 co-catalyst or both of H2SO4 and LiBr, though
the yields were decreased (3: 50% and 45%, respectively). Under
these conditions, N-cyclohexylidenemethylacetamide (4) was
obtained as a by-product in moderate yields (ca. 20%). It was
suggested that HCl was formed from PtCl2 under the reaction
conditions. Accordingly, the effect of HCl as well as platinum
sources on the reaction was further studied. In the absence of any
acids, Pt(PPh3)4 did not catalyze the reaction, while in the
addition of anhydrous HCl/dioxane, Pt(PPh3)4 exhibited the
catalytic activity (3: 40%, 4: 25%). On the other hand, other acids,
TFA and TfOH, were ineffective regardless of the acid strength.
In these cases, only formation of enamide 4 was observed (43–
45%). Similar acid effect was observed when Pt(dba)2 was used.
Interestingly, (R)-MOP served as an effective ligand in combina-
tion with Pt(dba)2 (3: 35%, 4: 17%), while PtCl2/MOP system
resulted in no formation of the desired product (vide supra).
In sharper contrast to PtCl2/phosphine system, K2PtCl4/
phosphine system showed higher catalytic activity when (R)-
MOP was employed as a ligand instead of biphPt-Bu2 (Table 1,
run 2 vs. 4 or run 3 vs. 5). Combinations of K2PtCl4 and
triarylphosphines also afforded the desired adduct in good yields
(runs 6 and 7). The coordination geometry (cis vs. trans) of the
plausible platinum halide phosphine complex might attribute to
this phenomena. Disappointingly, however, no chiral induction
of aldehydes with amides and carbon monoxide is described. In
contrast to precedent palladium catalysis, a remarkable ligand
acceleration by phosphines was observed. Furthermore, an
optically active N-acetyl amino acid was partially epimerized
under the platinum-catalyzed conditions, while faster racemiza-
tion was observed under the palladium catalysis.
The transition metal-catalyzed three-component coupling
reaction of aldehydes, amides, and carbon monoxide, so-called
amidocarbonylation or Wakamatsu reaction, is a versatile one-pot
method for the preparation of N-acylated a-amino acids.1 Cobalt-
catalysts were first discovered by Wakamatsu et al. in early
1970’s2 and palladium-catalysts were recently revealed by Beller
et al.3 The latter group also reported the catalytic activities of
rhodium, iridium, and ruthenium complexes in a patent.4 Among
these transition metals, the most active metal so far is palladium,
and now TON reaches 60000 in a certain case. However, despite
the importance of optically active a-amino acid derivatives,
asymmetric amidocarbonylation (either diastereoselective or
enantioselective) has not been accomplished yet.5 During the
course of our study to address this issue, we have found that
platinum catalysts were also effective for the reaction. Herein, we
describe preliminary results of our study on the platinum-
catalyzed amidocarbonylation.
Initially, we have examined the catalytic activity of various
transition metals for the reaction of cyclohexanecarboxaldehyde
(1) with acetamide (2) under the Beller’s conditions [LiBr
ꢁ
(35 mol%), conc. H2SO4 (cat.), and CO (60atm) at 120 C in N-
methylpyrrolidone (NMP)]. After many trials, an interesting
ligand effect was observed when platinum complexes were
employed (Eq 1).6 While PtCl2(PPh3)2 showed a moderate
catalytic activity (3: 15%), this was vanished in the presence of
DPPE. A Pt(0) complex, Pt(PPh3)4, exhibited similar activity to
PtCl2(PPh3)2 (3: 13%). On the other hand, the absence of any
phosphine ligand, that is, use of PtCl2(cod) resulted in loss of the
activity (3: trace), while the activity was dramatically enhanced
on the addition of 2-(di-tert-butylphosphino)biphenyl (biphPt-
Bu2)7 (3: 67%). In good contrast, the phosphorous ligand is not
essential for the catalytic activity of palladium catalysis. It has
been reported that palladium on carbon is the most effective
catalyst in palladium-catalyzed amidocarbonylation.8
CHO
Pt-catalyst
COOH
(5 mol%)
1
co-catalysts
NHAc
NHAc
(1)
+
+
CO (60 atm)
NMP, 120 °C, 15 h
AcNH2
3
4
2
Thus, we have examined various combinations of platinum
sources and phosphine ligands for the reactions of 1 with 2 under
Copyright ꢀ 2003 The Chemical Society of Japan

Chemistry Letters Vol.32, No.2 (2003)
161
Table 1. Effect of K2PtCl4/Phosphine/Acid catalystsa
In summary, we have disclosed that platinum/phosphine
catalysts exhibited a good catalytic activity for amidocarbonyla-
tion, and that a remarkable ligand acceleration by phosphines was
observed. It was also revealed that the undesired racemization of
the product, which would be a major obstacle for asymmetric
synthesis, could be retarded under the conditions. Further
investigations on the asymmetric amidocarbonylation are now
under way.
Conditions
Phosphine
Yield/%
Run
Acid
3
4
b
1
2
3
4
5
6
7
—
—
—
HCl
—
HCl
HCl
HCl
0
nd
biphPt-Bu2
biphPt-Bu2
(R)-MOP
(R)-MOP
PPh3
15
28
56
77
63
63
37
24
20
4
13
2
P(o-MeC6H3)3
aAll reactions were performed using 1 (1 mmol) and 2 (1 mmol) in the
presence of K2PtCl4 (5 mol%), a ligand (10mol%), and an acid
(10mol%) in NMP under CO (60atm) at 120 ꢁC for 15 h. bnd = not
determined.
This work was partially supported by New Energy and
Industrial Technology Development Organization (NEDO),
Japan Chemical Innovation Institute (JCII), SORST and CREST,
Japan Science and Technology Corporation (JST).
was observed when (R)-MOP was used as a chiral ligand.
Enamide 4 is not a dead end but one of the intermediates of
the platinum-catalyzed reaction. Thus, K2PtCl4/(R)-MOP or
trans-PtCl2(CH3CN)2/(R)-MOP (5 mol%/10mol%) was found to
catalyze carbonylation of 4 with water (1 equiv.) to give the
desired product 3 even in the absence of the acid co-catalyst (3:
64% and 65%, respectively).11
References and Notes
1
2
3
4
5
For an excellent review, see: M. Beller and M. Eckert, Angew. Chem.,
Int. Ed., 39, 1010 (2000).
H. Wakamatsu, J. Uda, and N. Yamakami, J. Chem. Soc., Chem.
Commun., 1971, 1540.
M. Beller, M. Eckert, F. Vollmuller, S. Bogdanovic, and H. Geissler,
¨
Angew. Chem., Int. Ed., 36, 1494 (1997).
K. Drauz, O. Burkhardt, M. Beller, and M. Eckert, DE 100 12 251 A1
(1999).
Furthermore, racemization experiments of (S)-N-acetylphe-
nylalanine (5) (98% ee) were performed under either platinum- or
palladium-catalyzed conditions [under CO (60atm) in NMP at
120 ꢁC for 15 h]. Almost no racemization of 5 was observed by
treatment of K2PtCl4/(R)-MOP catalyst (recovery of 5: 99%, 96%
ee), while PdBr2(PPh3)2/LiBr/H2SO4 catalyst3 promoted the
complete racemization (recovery of 5: 89%, 0% ee).12 In order to
ensure the formation of the catalytically active species, racemi-
zation experiments of 5 were also performed in the presence of 1
and 2 (5/1/2 = 0.5/1/1). With K2PtCl4/PPh3 catalyst (5 mol%/
10mol%), partial racemization of 5 was observed (recovery of 5:
quant., 80% ee) as well as formation of amino acid 3 (35%) and
enamide 4 (30%). On the other hand, PdBr2(PPh3)2 (5 mol%)
without co-catalysts resulted in faster racemization of 5 (recovery
of 5: 87%, 28% ee), though a higher catalytic activity was
observed (3: 79%, 4: 5%). Towards asymmetric catalysis of
amidocarbonylation, this racemization problem is one of the
major obstacles to be overcome. The milder conditions of the
platinum catalysis could contribute to this issue.
It was described in a review (ref. 1) that a low enantioselectivity (10%
ee) was attained by using 1-diphenylphosphanylethylbenzene as a chiral
ligand in palladium-catalyzed amidocarbonylation.
6
To avoid the contamination of palladium as possible, the platinum
complexes or salts with high purity were employed in this study:
PtX2(cod) (Strem, X = Cl and I, 99%; X = Br, 98%), PtCl2 (Strem,
99.9%, Pd 5 ppm%), Pt(PPh3)4 (Aldrich, 97%), K2PtCl4 (Aldrich,
99.99%), PtCl2(CH3CN)2 and Pt(dba)2 (prepared from K2PtCl4).
A. Aranyos, D. W. Old, A. Kiyomori, J. P. Wolfe, J. P. Sadighi, and S. L.
Buchwald, J. Am. Chem. Soc., 121, 4369 (1999).
M. Beller, W. A. Moradi, M. Eckert, and H. Neumann, Tetrahedron
Lett., 40, 4523 (1999).
J. P. Wolfe, R. A. Singer, B. H. Yang, and S. L. Buchwald, J. Am. Chem.
Soc., 121, 9550(1999).
7
8
9
10a) Y. Uozumi and T. Hayashi, J. Am. Chem. Soc., 113, 9887 (1991). b)
For a review, see: T. Hayashi, Acc. Chem. Res., 33, 354 (2000).
11 The intermediary of enamides in palladium-catalyzed amidocarbonyla-
tion was reported, see: D. A. Freed and M. C. Kozolwski, Tetrahedron
Lett., 42, 3403 (2001).
12 It was confirmed that an acid such as H2SO4 itself promoted the
racemization under the reaction conditions. On the other hand, Beller et
al. reported that rhodium or palladium complexes catalyzed the
racemization of N-acyl a-amino acids, see: a) M. J. Hateley, D. A.
Schichl, H. Kreusfeld, and M. Beller, Tetrahedron Lett., 41, 3821
(2000). b) M. J. Hateley, D. A. Schichl, C. Fischer, and M. Beller,
Synlett, 2001, 25.
13 General experimental procedure: In a 50mL stainless steel autoclave
with a glass liner and a magnetic stirring bar, a mixture of an aldehyde
(5 mmol), an amide (5 mmol), K2PtCl4 (5 mol%), triphenylphosphine
(10mol%), and HCl/dioxane (4 M, 10mol%) in N-methylpyrrolidone
(NMP) (5 mL) was heated at 120 ꢁC for 15 h under carbon monoxide (an
initial pressure: 60atm). The reaction was terminated by cooling to rt.
After releasing carbon monoxide, the deep green reaction mixture was
collected with NMP and analyzed by a reverse phase HPLC using 2,6-
dimethylphenol as an internal standard (YMC-pack ODS-A, 250 ꢂ
4:6 mm I.D.; CH3CN/H2O = 3/7 phosphate buffer solution). After
removal of NMP at 80–100 ꢁC under a reduced pressure, dichlorome-
thane (30mL) and sat. aqueous NaHCO 3 (30mL) were added to the
residue. The mixture was stirred well and filtered through a Celite pad.
The aqueous layer was separated, washed with dichloromethane
(3 ꢂ 30 mL), acidified with 85% phosphoric acid to ca. pH 2, and
extracted with ethyl acetate (3 ꢂ 30 mL). The combined organic layers
were dried over anhydrous MgSO4, filtered, and concentrated in vacuo.
The crystalline residue was washed with diethyl ether, collected by
filtration, and dried under vacuum to give an analytically pure N-acyl
amino acid.
Finally, substrate scope of the platinum-catalyzed reaction
was surveyed (Table 2). All reactions were performed in a
5 mmol scale of aldehydeꢁs using K2PtCl4/PPh3/HCl catalyst
under CO (60atm) at 120 C for 15 h.13 It was found that the
desired amidocarbonylation products were obtained in moderate
to good yields, though reaction conditions were not optimized yet.
Table 2. Pt-catalyzed amidocarbonylation (see Ref. 13)
Run
R1
R2
Yield/%a
1
2
3
4
5
6
c-C6H11
c-C6H11
PhCH2CH2
Me2CHCH2
i-Pr
Me
Ph
Me
Me
Me
Me
53 (70)
60(69)
28 (29)
44 (47)
46
Ph
32
aIsolated yields. Yields determined by HPLC analyses are in parenth-
eses.