A practical procedure for the synthesis of multifunctional aldehydes through the Fukuyama reduction and elucidation of the reaction site and mechanism

Article abstract of DOI:10.1016/j.tetlet.2004.02.130

A highly efficient heterogeneous Pd/C catalyst D1 was found to effect the reduction of thiol esters 1 to the corresponding aldehydes 2 with such a low catalyst loading as 0.5-1.0mol%. The chemical properties of the Pd/C catalysts together with the XRF analysis reveal that the reduction is most likely to proceed on the solid surface of the Pd/C catalyst rather than in the solution phase outside the pores. A reaction mechanism through oxidative addition of Pd to the thiol esters 1 was postulated by detection of the oxidative addition intermediate by React IR analysis. A practical purification of 2 was accomplished by conversion to water-soluble bisulfite adducts 7.

Full text of DOI:10.1016/j.tetlet.2004.02.130

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 3219–3223
A practical procedure for the synthesis of multifunctional
aldehydes through the Fukuyama reduction and elucidation
of the reaction site and mechanism
Mayumi Kimura and Masahiko Seki*
Process Chemistry Research Laboratories, Tanabe Seiyaku Co., Ltd, 3-16-89, Kashima, Yodogawa-ku, Osaka 532-8505, Japan
Received 6 February 2004; revised 24 February 2004; accepted 26 February 2004
Abstract—A highly efficient heterogeneous Pd/C catalyst D1 was found to effect the reduction of thiolesters 1 to the corresponding
aldehydes 2 with such a low catalyst loading as 0.5–1.0 mol %. The chemical properties of the Pd/C catalysts together with the XRF
analysis reveal that the reduction is most likely to proceed on the solid surface of the Pd/C catalyst rather than in the solution phase
outside the pores. A reaction mechanism through oxidative addition of Pd to the thiolesters 1 was postulated by detection of the
oxidative addition intermediate by React IR analysis. A practical purification of 2 was accomplished by conversion to water-soluble
bisulfite adducts 7.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Multifunctional aldehydes have received considerable
attention as usefulintermediates for drugs and natural
products. Various synthetic methods1 of the aldehydes
involving reduction of activated or nonactivated carb-
oxylic acid derivatives have been reported. However,
they generally suffer from either one of the following
drawbacks: functionalgroup intoelrance, the need for
quite low temperature and/or expensive reagents. A two-
step procedure involving prior reduction of carboxylic
acids to alcohols followed by mild oxidation has thus
been the most reliable and versatile method to prepare
aldehydes. Fukuyama and co-workers have recently
developed an elegant synthesis of the aldehydes through
by way of the bisulfite adducts. A possible reaction site
and mechanism were discussed as well.
The Fukuyama reduction was optimized with thiolester
1a3 used as a typicalsubstrate (Eq. 1). In our initial
study, the reaction was tested using Et3SiH (2 equiv) and
the Pd/C catalyst D34 (Pd: 10 wt %, 0.5 mol% relative to
1a, H2O: 1.8 wt %) in various solvents (Fig. 1).
Although, in the reported procedure,2 acetone or
CH2Cl2 has been employed as the reaction solvent, a
higher reaction rate and a conversion were observed in
THF. While acetone may have an advantage of trapping
reduction of thiolesters with Et SiH in the presence of
3
Pd/C catalyst.2 The reaction takes place under mild
reaction conditions, that is, at room temperature, for the
substrates carrying various functionalgroups. It is,
however, not entirely satisfactory for a practical use
because of the need for relatively high catalyst loading
(ca. 5 mol%) and purification of the products by silica-
gelcoulmn chromatography. We describe herein a
highly efficient Pd/C catalyst that allows the Fukuyama
reduction with an extremely low catalyst loading (0.5–
1.0 mol%), and a practical purification of the products
100
THF
80
acetone
60
CH2Cl2
40
20
0
2-propanol
0
5
10
Period [h]
15
20
Figure 1. Influence of the solvent on the Fukuyama reduction of 1a to
2a. The reaction was conducted by the use of Et3SiH (2.0 equiv) and
Pd/C catalyst D3 (0.5 mol%).
* Corresponding author. Tel.: +81-6-6300-2456; fax: +81-6-6300-2816;
e-mail: m-seki@tanabe.co.jp
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.02.130

3220
M. Kimura, M. Seki / Tetrahedron Letters 45 (2004) 3219–3223
O
O
100
80
60
40
20
0
Et3SiH
SR2
H
D1
D2
D3
R1
R1
Pd(0)
O
Pd
Pd
Pd(0)
Et3SiSR2
3
4
path A
D4
D5
O
R1
SR2
R1
H
D6
1
2
path B
OSiEt3
SR2
OSiEt3
W2
Et3SiSR2
R1
R1
SR2
Et3SiPdH
Pd
0
5
10
Period [h]
15
20
H
Pd(0)
Pd(0)
Et3SiH
6
5
Figure 3. Variation of the Pd/C catalyst in the Fukuyama reduction of
1a to 2a. The reaction was conducted by the use of Et3SiH (2 equiv)
and Pd/C catalyst (0.5 mol %) in THF.
Scheme 1. Possible mechanisms for the Fukuyama reduction.
Et3SiSiEt generated in the reaction to prevent its further
reaction with aldehydes, that is formation of O-trieth-
ylsilyl-O,S-acetals (Scheme 1, compound 6), the use of
THF appeared not to be accompanied by such side
reactions.
reflect that diffusion of 1a to the Pd surface of the cat-
alyst is a significant factor for the successful reaction.
The catalyst D1 is thus considered to be the most
effective in the reduction of 1a to provide aldehyde 2a in
92% isolated yield with such a low catalyst loading as
0.5 mol% (Table 2, entry 1).
Pd/C D3
CHO
COSEt
While the oxidized catalysts D4, D5, and D6 required
long periods for the onset of the reaction than D1, D2,
and D3, they showed the similar order of the activities
with respect to Pd impregnation depth: D4 (thick
shell) > D5 (egg shell) > D6 (uniform). Presence of water
in the catalyst is detrimental for the reaction. The cat-
alyst W2 whose water content is 58% showed no cata-
lytic activity. It is thus concluded that the most effective
catalyst D1 is characterized by thick shell type Pd dis-
tribution, a low oxidation state, a high Pd dispersion
and a low water content (Table 1).
(0.5 mol%)
Et3SiH (2 equiv)
r.t.
MeO
MeO
2a
1a
ð1Þ
Screen of the Pd/C catalysts D1, D2, D4, D5, D6, and
W24 (Pd: 5 wt %, 0.5 mol% relative to 1a) as well as the
catalyst D3, having various chemicalproperties, in the
reduction of 1a was the next subject for our investiga-
tion.5 The Pd distribution may have a significant impact
on the catalytic efficiency. The Pd catalysts tested in this
study have three types of Pd distribution: egg shell, thick
shell, and uniform (Fig. 2). In the egg shell catalyst, Pd
distributes close to the surface within 50–150 nm depth.
In contrast, Pd disperses homogeneously in the uniform
catalyst. The thick shell catalyst denotes the one whose
Pd distributes until200–500 nm from the surface.
To investigate the reaction mechanism of the Fukuyama
reduction, React IR analysis of the reaction mixture was
conducted (Fig. 4).7 By scanning the IR spectrum once
for every 1 min from 650–4000 cmÀ1 during the reaction,
three components that changed the concentration were
Table 1. Chemicalproperties of the Pd/C catayl sts
As shown in Figure 3, the reduced catalyst D1, D2, and
D3 exerted higher reaction rates than the oxidized
counterparts D4, D5, and D6. The highest initialacce-l
eration of the reaction was observed in the catalyst D2
(egg shell) whose Pd is distributed closer to the surface
than D1 (thick shell). However, the reaction almost
stopped in 1 h with a conversion of 75%. Although the
increased surface area in the egg shell catalyst D2 would
enhance the reaction, it might lead to a facile precipi-
tation of Pd on the solid support to retard the reaction.6
The lower activity found in uniform catalyst D3 may
Pd/C
Pd distribu- Impreg- Reduc-
Pd dis- Water
persion content
catalysta tionb
nation
depth
(nm)b
tion
degree
(%)c
(%)d
(wt %)
D1
D2
W2
D3e
D4
D5
D6
Thick shell
200–500 25–99
26
27
28
36
35
28
36
3.0
4.2
Egg shell
Egg shell
Uniform
Thick shell
Egg shell
Uniform
50–150
50–150
––f
25–99
25–99
25–99
58
1.8
200–500 0–25
1–1.5
3.2
1–1.5
50–150
––f
0–25
0–25
a Purchased from Degussa Japan Co., Ltd, Catalysts Division except
D3.
b Estimated by transmission electron microscopy (TEM).
c Detected by temperature programmed reduction (TPR) measure-
ment.
d Pd dispersion is calculated with CO chemisorption.
e Purchased from Kawaken Fine Chemicals.
f Not determined.
50-150 nm
200-500 nm
Pd
.
.
. .
.. .
.
.
.
.
..
.
.
.
..
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. .
.
.
.
.
.
.
.
.
.
.
. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Egg shell
Thick shell
Uniform
Figure 2. Distribution of Pd in the Pd/C catalysts.

M. Kimura, M. Seki / Tetrahedron Letters 45 (2004) 3219–3223
Table 2. The Fukuyama reduction of thiolesters 1a–f
3221
O
O
Et3SiH
R1
H
R1
SEt
Pd/C catalyst D1
2a-f
1a-f
THF, r.t.
Entry
1
Thiolester 1
MeO
Catalyst D1 (mol%)
Et 3SiH (equiv)
2
t (h)
Aldehyde 2
Yield (%)a
92
COSEt
0.5
2
2a
1a
1b
COSEt
2
3
0.5
0.5
2
2
2
5
2b
2c
93
90
MeO
MeO
COSEt
1c
CO2Ph
N
4
5
1.0
1.0
3
3
17
17
2db
92
97
Ph
COSEt
S
1d
O
BnN
NBn
2e
COSEt
S
1e
Bn
2fb
90
N
6
1.0
3
17
O
COSEt
S
1f
a Isolated yield after purification with silica-gel column chromatography.
b Opticalpurity of 2d and 2f was confirmed to be >99% ee.
detected. Comparing with the IR spectra of the
authentic samples, the blue and black lines are assigned
as thiolester 1a and aldehyde 2a, respectively. Another
species, shown in the red line, is of particular interest. It
showed an absorbance at 1692 cmÀ1 and initially grew,
but then slowly disappeared as the reaction progressed.
This absorbance can be attributed to transient form-
ation of an intermediate carrying a carbonylgroup at-
tached to Pd (Scheme 1, compound 3 or 4).8
1692 cm-1
1690 cm-1
1724 cm-1
1650
(a)
1700
1750
In the carbon–carbon bond forming reactions between
thiolesters and boronic acids, oxidative addition of thiol
esters to Pd(0) has been suggested as a possible reaction
mechanism because of the high thiophilicity of Pd.9 In
the present reaction, path A involving the oxidative
addition of thiolester 1 to Pd and subsequent trans-
metallation with triethylsilane may be operative as well
(Scheme 1). Although an alternative reaction mecha-
nism, path B, which takes place through direct hydro-
silylation of the carbonyl group of 1 may be possible,10 it
should be ruled out since compound 3 or 4 does exist
(vide supra) and any O-triethylsilyl-O,S-acetals 6 were
not detected at all in the reaction mixture.
Wavenumber (cm-1)
0
50
100
Period [min]
150
(b)
To elucidate the reaction site as well as the loss of Pd
from the solid support, the concentration of Pd in the
solution during the reaction was measured by XRF (X-
ray fluorescence) analysis of the reaction mixture peri-
odically obtained and filtered through Celite. Only a
Figure 4. React IR analysis of the Fukuyama reduction of 1a to 2a: (a)
component spectrum, (b) component profile. Curves are shown for 1a
(
by the use of Et3SiH (2 equiv) and Pd/C catalyst D1 (0.5 mol%) in
), 2a (
), and an intermediate (
). The reaction was conducted
THF.

3222
M. Kimura, M. Seki / Tetrahedron Letters 45 (2004) 3219–3223
small amount of Pd was found to bleed from the solid
support during the reaction though a maximum con-
centration of Pd (4.5% relative to the initially added
catalyst D1 or 0.023 mol % relative to the initially added
1a) was found in 0.5 h. The reduction did not proceed to
any appreciable extent in the presence of such a small
amount of Pd catalyst [0.023 mol % of the catalyst D3 or
Pd(PPh3)4]. The pore size distribution plays a significant
role for the activity of the Pd/C catalyst as well.11 The
Pd/C D1 possesses pore volumes of 0.9–1.3 mL/g and ca.
50% of the pores consists of mesopores (diameters: 2–
50 nm) where most of the organic molecules involving
thiolesters 1, aldehydes 2, and the Pd intermediates 3
and 4 can penetrate through the pores.11
The Fukuyama reduction of thiolesters 1b–f carrying
various functionalgroups was tested using the present
protocolto provide the corresponding adl ehydes 2b–f in
high yields in the presence of 0.5–1.0 mol % of the cat-
alyst D1 (Table 2). Although the compound 2f, an
intermediate for (+)-biotin,13 is quite labile for racemi-
zation, the present reaction took place with the stereo-
genic center retained as such, which was confirmed by
chiral HPLC analysis of the alcohol obtained by
reduction of 2f with NaBH4.
Although the Fukuyama reduction generally affords
aldehydes in excellent yields, it is often difficult to sep-
arate the product from unreacted Et3SiH, Et3S:SEt and/
or Et3SiSiEt3.14 The problem becomes much more seri-
ous when thiolester 1g obtained from an odorless high
boiling point dodecane thiol15 is employed as the sub-
strate. We envisioned a possible use of a water-soluble
bisulfite adduct 7 for the purification of the product
(Scheme 2). After the reduction of 1a, 1f, and 1g, the
reaction mixture was filtered. Into the filtrate was added
an aqueous solution of sodium bisulfite and the mixture
was vigorously stirred. In the case of 1a and 1g, the
bisulfite adduct 7a was obtained in 86% and 90% yield,
respectively, as water-soluble crystalline solids by simple
filtration. While the bisulfite adduct 7f did not crystallize
without concentration, it was obtained quantitatively as
an aqueous solution by simple separation of the aqueous
phase. Treatment of the adduct 7a and 7f (aqueous
solution) with aq K2CO3 in ether quantitatively pro-
vided pure aldehyde 2a and 2f, respectively.16
These results together with the chemical properties of
the optimized Pd/C catalyst D1 may suggest that the
reaction takes place predominantly on the Pd surface of
the Pd/C catalyst not in the solution phase outside the
pores.12 The reactant 1 should adsorb on the Pd surface
in the charcoal matrix of the Pd/C catalyst while the
product 2 being continuously pushed out from there
(Fig. 5). This is in remarkable contrast with Heck
reactions6 that have been supposed to be catalyzed by
Pd dissolved in solution. After the reaction, virtually
complete recovery of Pd (>95%) was accomplished by
simple filtration and subsequent treatment involving
combustion of the solids obtained.
O
O
+ Et3SiSEt
+ Et3SiH
SEt
R
R
H
1
2
In conclusion, a practical synthesis of multifunctional
aldehydes through Fukuyama reduction was worked
out by combining the use of readily accessible Pd/C
catalyst D1 and the simple purification via bisulfite
adducts. The reduction may take place on the surface of
the Pd/C catalyst not in the solution phase outside the
pores. The considerable low catalyst loading as well as
the simple work-up of the present protocol would permit
a facile access to the multifunctional aldehydes, com-
pounds of glowing significance.
Solution Phase
Activated Charcoal
2 - 50 nm
200 - 500 nm
Pd
Pd
Pd
Figure 5. A possible reaction site for the Fukuyama reduction.
OH
O
a
R1
SR2
SO3Na
MeO
1a: R1 = 4-MeO-Ph-(CH2)2-, R2 = Et
7a
Bn
1g: R1 = 4-M eO-Ph-(CH) -, R2 = CH3(CH2)11
-
2 2
O
N
Bn
O
N
OH
S
R2 = Et
1f: R1
=
S
SO3Na
7f
O
b
R1
2a: R1 = 4-MeO-Ph-(CH2)2-
Bn
H
O
N
2f: R1
=
S
Scheme 2. Isolation of aldehydes 2 through bisulfite adducts 7. (a) (i) Pd/C, Et3SiH, THF, rt, 2 h, 90%, (ii) NaHSO3, AcOEt, H2O, 86% (from 1a),
quant. (from 1f), 90% (from 1g); (b) K2CO3, H2O, Et2O, quant.

M. Kimura, M. Seki / Tetrahedron Letters 45 (2004) 3219–3223
3223
9. (a) Savarin, C.; Srogl, J.; Liebeskind, L. S. Org. Lett. 2000,
2, 3229; (b) Liebeskind, L. S.; Srogl, J. J. Am. Chem. Soc.
2000, 122, 11260.
Acknowledgements
The authors wish to thank Dr. Kiyoaki Sugi, Degussa
Japan Co., Ltd, Catalysts Division for useful discussions
concerning this work. The authors acknowledge De-
gussa Japan Co., Ltd, for disclosing the chemical
properties of the Pd/C catalysts D1–6 and W2. Thanks
are also extended to Mr. Tsuyoshi Furukawa, S.T.
Japan Inc. for the usefulcomments and discussion on
the React IR analysis.
10. (a) Ojima, I.; Nihonyanagi, M.; Nagai, Y. Chem. Com-
mun. 1972, 938; (b) Ojima, I.; Nihonyanagi, M.; Nagai, Y.
Bull. Chem. Soc. Jpn. 1972, 45, 3722; (c) Semmelhack, M.
F.; Misra, R. N. J. Org. Chem. 1982, 47, 2469; (d) For
review, see: Brunner, H. Angew. Chem., Int. Ed. Engl.
1983, 22, 897; (e) For review, see also: Ojima, I. The
Hydrosilylation Reaction. In The Chemistry of Organic
Silicon Compounds; Patai, S., Rappoport, Z., Eds.; Wiley:
Chichester, UK, 1989; Vol. 1, Chapter 25, p 1479; (f)
Felfoldi, K.; Kapocsi, I.; Bartok, M. J. Organomet. Chem.
1989, 362, 411.
References and notes
11. Auer, E.; Freund, A.; Pietsch, J.; Tacke, T. Appl. Catal. A:
General 1998, 173, 259.
1. Larock, R. C. Comprehensive Organic Transformations;
Jone Wiley: New York, 1999; pp 1234–1236, 1238–1247,
1263–1272 and 1474.
2. (a) Fukuyama, T.; Lin, S.-C.; Li, L. J. Am. Chem. Soc.
1990, 112, 7050; (b) Kanda, Y.; Fukuyama, T. J. Am.
Chem. Soc. 1993, 115, 8451; (c) Fujiwara, A.; Kan, T.;
Fukuyama, T. Synlett 2000, 1667; (d) Tokuyama, H.;
Yokoshima, S.; Lin, S.-C.; Li, L.; Fukuyama, T. Synthesis
2002, 1121.
3. Seki, M.; Kondo, T.; Iwasaki, T. J. Chem. Soc., Perkin
Trans. 1 1996, 2851.
4. The Pd/C catalyst D1, D2, D4–D6, W2 (Degussa Japan
Co., Ltd, Catalysts Division) and D3 (Kawaken Fine
Chemicals) are commercially available.
12. The reaction site for the Ni/C-catalyzed Suzuki, Kumada,
Negishi, amination and dehydrohalogenation reactions,
see: Lipshutz, B. H.; Tasler, S.; Chrisman, W.; Spliethoff,
B.; Tesche, B. J. Org. Chem. 2003, 68, 1177, and references
cited therein.
13. Seki, M.; Kimura, M.; Hatsuda, M.; Yoshida, S.; Shimizu,
T. Tetrahedron Lett. 2003, 44, 8905.
14. One of the side reactions of the present synthesis:
2Et3SiH fi H2+Et3SiSiEt3, see: Lautens, M.; Kumanovic,
S.; Meyer, C. Angew. Chem., Int. Ed. Engl. 1996, 35, 1329.
15. Miyazaki, T.; Han-ya, Y.; Tokuyama, H.; Fukuyama, T.
Synlett 2004, 477.
16. A typicalprocedure for the Fukuyama reduction (prep-
aration of 2a). To a solution of 1a (10 g, 45 mmol) in THF
(45 mL) was added Pd/C catalyst D14 (5 wt %, 3 wt %
water, 474 mg, 0.22 mmol) under N2 atmosphere. To the
suspension was added dropwise Et3SiH (14 mL, 89 mmol)
over 1 h. After stirring the mixture at 20 ꢁC for 1 h, the
mixture was filtered. The filtrate was evaporated and to
the residue were added AcOEt (30 mL), H2O (30 mL) and
NaHSO3 (5.4 g, 52 mmol) and the mixture was vigorously
stirred at 20 ꢁC for 17 h. The solids formed were collected,
washed with AcOEt and dried to give compound 7a (10 g,
86%) as colorless crystals. Mp 165.6–166.5 ꢁC; IR (ATR)
5. An example where the chemical properties of the Pd/C
catalyst affect the reaction, see: Sajiki, H.; Ikawa, T.;
Hirota, K. Tetrahedron Lett. 2003, 44, 7407.
6. (a) Heidenreich, R. G.; Krauter, J. G. E.; Pietsch, J.;
€
Kohler, K. J. Mol. Catal. A 2002, 182–183, 499; (b)
€
Kohler, K.; Heidenreich, R. G.; Krauter, J. G. E.;
Pietsch, J. Chem. Eur. J. 2002, 8, 622.
7. The Fukuyama reduction monitored by React IR. React
IR analyses were performed on Mettler Toledo React IR
4000. The settings are as follows: detector, MCT; probe
type, Di-Comp; scans, 8; X-axis max, 4000 cmÀ1; X-axis
min, 650 cmÀ1; resolution, 4. The reaction was conducted
on thiolester 1a (10 g, 44.6 mmol), Et3SiH (14.2 mL,
88.9 mmol), Pd/C D1 (5 wt %, 3 wt % water, 474 mg,
0.22 mmol) and THF (200 mL) using four necked flask
(500 mL) fitted with the React IR probe. The reaction was
monitored for 2.5 h.
m ¼ 1611, 1515, 1033, 635 cmÀ1
;
1H NMR (DMSO-d6)
d 7.09 (d, J ¼ 8:0 Hz, 2H), 6.83 (d, J ¼ 8:0 Hz, 2H), 5.36
(d, J ¼ 4:0 Hz, 1H) 3.34–3.81 (m, 4H), 2.50 (s, 3H); SIMS
m=z 245 (Mþ+1); Anal. Calcd for C10H13O5SNa: C, 44.77;
H, 4.88. Found: C, 44.37; H, 4.76. To the solution of
K2CO3 (25 g, 0.18 mol) in water (175 mL) was added the
compound 7a (10 g, 37 mmol) and Et2O (75 mL). After
stirring the mixture at 20 ꢁC for 2 h, the organic phase was
separated, washed twice with water and dried over
anhydrous MgSO4 and evaporated to afford 2a (6.1 g,
quant) as colorless oil. IR (ATR) m ¼ 1720, 1611 cmÀ1; 1H
NMR (CDCl3) d 9.77 (s, 1H), 7.09 (d, J ¼ 8 Hz, 2H), 6.82
(d, J ¼ 8 Hz, 2H), 3.78 (s, 3H), 2.70–2.90 (m, 4H); SIMS
m=z 164 (Mþ+1).
8. Ito, T.; Tsuchiya, H.; Yamamoto, A. Bull. Chem. Soc. Jpn.
1977, 50, 1319.
O
1676 cm-1(νC=O
Me
)
Me3P
O
Pd
PMe3
Me
O