Preparation of acyloins by cerium-catalyzed, direct hydroxylation of β-dicarbonyl compounds with molecular oxygen

Article abstract of DOI:10.1002/ejoc.200390075

We report the metal-catalyzed α-hydroxylation of a variety of cyclic and acyclic β-dicarbonyl compounds by molecular oxygen. The decisive advantage of this new method is the use of catalytic amounts of the nontoxic cerium salt CeCl₃·7H₂O in 2-propanol at ambient temperature. Most of the cyclic substrates 4a-4i give high yields of analytically pure products 5a-5i, and the workup procedure is simple filtration through silica gel. The oxidation of acyclic dicarbonyl compounds 4j-4p, however, is accompanied by side reactions and decomposition, reducing the yields of products 5j-5p significantly. A proposed mechanism is in agreement with experimental results, in particular the observed oxygen uptake. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200390075

FULL PAPER
Preparation of Acyloins by Cerium-Catalyzed, Direct Hydroxylation of
β-Dicarbonyl Compounds with Molecular Oxygen
Jens Christoffers,*[a] Thomas Werner,[a] Sven Unger,[a] and Wolfgang Frey[a]
Keywords: Catalysis / Cerium / β-Dicarbonyl compounds / Oxidation / Oxygen
We report the metal-catalyzed α-hydroxylation of a variety
of cyclic and acyclic β-dicarbonyl compounds by molecular
oxygen. The decisive advantage of this new method is the
use of catalytic amounts of the nontoxic cerium salt
CeCl3·7H2O in 2-propanol at ambient temperature. Most of
the cyclic substrates 4a−4i give high yields of analytically
pure products 5a−5i, and the workup procedure is simple fil-
compounds 4j−4p, however, is accompanied by side reac-
tions and decomposition, reducing the yields of products
5j−5p significantly. A proposed mechanism is in agreement
with experimental results, in particular the observed oxy-
gen uptake.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
tration through silica gel. The oxidation of acyclic dicarbonyl Germany, 2003)
Introduction
stoichiometric amount of strong base. Since acetone is the
only by-product formed in these DMD oxidations,[11] this
method may be viewed as environmentally sound ‘‘green
chemistry’’. The yields can be improved by the application
of either KF (supposedly stabilizing the enol tautomer by
H-bonding)[12] or catalytic amounts of [Ni(acac)2].[13] How-
ever, with regard to economical and ecological considera-
tions molecular oxygen represents the optimal oxidizing re-
agent, an issue that has received continuous attention.[14]
The metal-catalyzed α-hydroxylation of β-dicarbonyl com-
pounds with O2 has previously been reported three times.
CoCl2,[15] Cs2CO3,[16] and Mn(OAc)2·4H2O[17] have been
applied as catalytically active compounds. There has been
one publication on Mn- and Co-catalyzed radical additions
of malonates to olefins in the presence of oxygen yielding
γ-hydroxylated products.[18] In this paper we report the ap-
plication of CeCl3·7H2O as a catalyst[19] that seems to be
superior to the other protocols mentioned above. The ad-
vantages, apart from the improved yields, are the use of the
inexpensive and nontoxic cerium salt and iPrOH as the
solvent.
The α-hydroxy-β-dicarbonyl moiety can be found in bio-
logically important compounds such as the cyclopentenoid
kjellmanianone (1),[1] the biosynthetic precursors of valine
and isoleucine [α-acetolactate (2a) and α-acetohydroxybu-
tyrate (2b)],[2] and tetracycline-type antibiotics such as
doxycycline (3; Scheme 1).[3] In addition to some other
methods of synthesizing this functionality[4] a number of
reagents exist for the direct α-hydroxylation of β-dicarbonyl
compounds. Examples are the stoichiometric application of
[Pb(OAc)4],[5] the Mimoun reagent (MoO5·pyridine·
HMPA),[6] oxaziridines introduced by Davis[7] and perac-
ids.[8] A sequence of deprotonation, silyl enol ether forma-
tion and epoxidation known as Rubottom oxidation[9] can
also be applied. The disadvantage of all these methods is
the formation of large quantities of by-products. Adam and
co-workers were the first to use dimethyl dioxirane (DMD)
for the α-oxidation of ketones after deprotonation with
LDA or NaHMDS.[10] The corresponding α-hydroxylation
of β-dicarbonyl compounds with DMD did not require a
Results and Discussion
Optimization
In the course of our combinatorial type of search for new
catalysts for CϪC bond-forming reactions[20] we observed
the formation of α-hydroxylated by-products 5 when treat-
ing β-dicarbonyl compounds 4 with catalytic amounts of
CeCl3·7H2O without excluding air. During our search for
optimal reaction conditions we recognized that the solubil-
ity of oxygen in the reaction mixture seems to be crucial
for the success of this transformation. Finally, the optimal
solvents turned out to be DMF and iPrOH, and the conver-
Scheme 1. Examples of the α-hydroxy-β-dicarbonyl moiety occur-
ring in natural products
[a]
Institut für Organische Chemie der Universität Stuttgart,
Pfaffenwaldring 55, 70569 Stuttgart, Germany,
Fax: (internat.) ϩ 49-(0)711/685-4269
E-mail: jchr@po.uni-stuttgart.de
Eur. J. Org. Chem. 2003, 425Ϫ431  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ193X/03/0203Ϫ0425 $ 20.00ϩ.50/0
425

J. Christoffers, T. Werner, S. Unger, W. Frey
FULL PAPER
sion should be run at ambient temperature. Importantly, we products needed to be removed by either filtration through
found that 1 atm of O2 was in most cases superior to air. SiO2 (5d, 5e, 5f) or chromatography (5g). The six-membered
Nevertheless, full conversion of some substrates, for ex- ring products 5h and 5i seemed to decompose slowly under
ample 4a or 4c, can also be achieved with air. Moreover, a the reaction conditions and they had to be purified by dis-
slow but continuous stream of O2 seemed to be required for tillation or chromatography. In the case of 5h the acid 6
optimal results. The use of a static reservoir (e.g. a balloon (formation of 6 from 5h has been reported before)[8,17] was
of O2) sometimes gave irreproducible results. Presumably identified as the decomposition product (28%). The chloro
the slow evaporation of the solvent lowers the partial pres- compound 7a[21] was found as a by-product (4%).
sure of O2 adjacent to the liquid phase. All optimization
In the cases of the lactams 5d and 5e, as well as the capro-
experiments were performed with compounds 4a and 4h as lactone 5f, the constitution of the product was proved by
the substrates. With regard to the cerium source, we recom- X-ray single-crystal analysis. Their molecular structures are
mend CeCl3·7H2O, although (NH4)2Ce(NO3)6 or shown in Figures 1Ϫ3.
Ce(OAc)3·H2O can also be used in some cases. The use of
additives (Brönstedt bases or ligands) never improved the
conversion.
The amount of CeCl3·7H2O applied in our standard pro-
cedure was 5 mol %. However, it is worth mentioning that
for substrate 4a, which seems to be the optimal case, full
conversion was observed (by GC) within 16 h with 1 mol %
catalyst, and even with 0.5 mol % a 77% yield of isolated
material 5a was achieved.
Cyclic Substrates
With our optimized procedure, cyclic substrates 4aϪ4i
generally gave good (5dϪ5i) to excellent (5aϪ5c) results
(Scheme 2). Moreover, at least for compounds 5aϪ5g, the
Figure 1. Molecular structure of the α-hydroxylated butyrolactam
5d
workup procedure was extraordinarily simple: Filtration
through a small pad of SiO2 separated all metal-containing
materials and gave analytically pure products 5aϪ5g after
removal of all volatiles under high vacuum. The ester 5b
was obtained as a mixture of two diastereoisomers (1:1). In
the case of the lactams 5d, 5e and lactones 5f, 5g polar by-
Figure 2. Molecular structure of the α-hydroxylated valerolactam
5e
Scheme 2. Cerium-catalyzed α-hydroxylation of cyclic β-dicarbonyl
compounds 4aϪ4i; products 5aϪ5f were purified by simple filtra-
tion through 2 cm of SiO2, 5g was chromatographed, and 5h, 5i
were purified either by chromatography (38 and 46% yield for 5h
and 5i, respectively) or by Kugelrohr distillation; in the case of 5h
by-products 6 and 7a were isolated after chromatographic workup
Figure 3. Molecular structure of the α-hydroxylated caprolactone
5f
in 28 and 4% yield, respectively
426
Eur. J. Org. Chem. 2003, 425Ϫ431

Preparation of Acyloins from β-Dicarbonyl Compounds
FULL PAPER
Acyclic Substrates
ponding α-hydroxylated compounds 5vϪ5z were only de-
tectable in trace amounts by GC-MS.
Investigation of acyclic substrates showed that only the
β-oxo esters 4jϪ4o and the β-diketone 4p with an α-alkyl
substituent gave isolable α-hydroxylated products 5jϪ5p
(Scheme 3). As shown in Scheme 3, the yields (30Ϫ44%) are
moderate, in the case of the α-isopropyl ester 5m even low.
This low yield might be due to a steric effect. For two of
the substrates (4o, 4p), the α-chlorinated by-products 7b[22]
and 7c[19] were separated and isolated by column chromato-
graphy.
We utilized the enol acetate 8a as well as the silyl enol
ethers 8b, 8c as substrates hoping to find isolable interme-
diates of the oxidation. Unfortunately, we detected no con-
version for compounds 8aϪ8c, the allylic alcohol 8d or the
ether 8e.
Mechanistic Rationale
From a mechanistic point of view, we propose the in situ
oxidation of CeIII to CeIV by O2. According to Equation
(1) in Scheme 5, 2 equiv. of CeIII require 1 equiv. of 1/2
O2. We further assume the formation of CeIV diketonate
complexes A in the reaction mixture. Species A could un-
dergo intramolecular electron transfer (ligand-to-metal
charge transfer) giving a radical species B. At least formally,
species B is oxidized by a second equivalent of CeIV and
hydrated to give products 5. Although the details of the
conversion of B to 5 remain unclear, this proposal is in ac-
cordance with the stoichiometry in Equation (1), and it is
experimentally supported by the average oxygen uptake,
calculated from the mass difference to be 1.15 equiv. of 1/2
O2 per equivalent of dicarbonyl compound 4a, if 5 mol %
CeIII catalyst is used. Application of 10 mol % catalyst re-
sults in an uptake of 1.22 equiv. of oxygen and 50 mol %
catalyst in an uptake of 1.37 equiv. of oxygen per equivalent
of ester 4a. This stoichiometry per se excludes the oxidation
of any co-substrate, such as the solvent iPrOH, to give ace-
tone.
Scheme 3. Cerium-catalyzed α-hydroxylation of acyclic β-dicar-
bonyl compounds 4jϪ4p; products 5jϪ5p were purified by chroma-
tography; in the case of 5o and 5p by-products 7b and 7c were
isolated in 14 and 7%, respectively; starting materials 4m (63%), 4n
(16%) and 4o (42%) were recovered
Other Substrates
The diketones 4qϪ4s as well as the cyano ester 4t and
the sulfone 4u (Scheme 4) showed no conversion under the
reaction conditions. In contrast, the dicarbonyl compounds
4vϪ4z gave unspecified reaction mixtures, and the corres-
Scheme 5. Mechanistic proposal for the cerium-catalyzed α-oxida-
tion of β-dicarbonyl compounds
The outline shown in Scheme 5 is further supported by
the fact that an alkyl group (R ϶ H) seems to be required
for chemoselective conversion of the substrates, since the
electron-deficient intermediate B is stabilized by hypercon-
jugation. Moreover, the formation of chloro compounds
7aϪ7c as by-products can be rationalized from interme-
diate B.
Scheme 4. Substrates either giving no conversion (4qϪ4u, 8aϪ8e)
or decomposing under the reaction conditions (4vϪ4z)
Eur. J. Org. Chem. 2003, 425Ϫ431
427

J. Christoffers, T. Werner, S. Unger, W. Frey
FULL PAPER
(50.0 mmol, 25.0 mL, 2.0 ) in THF (25 mL). Subsequently, a solu-
tion of anhydrous MeOAc (7.98 mL, 7.41 g, 100 mmol) in THF (10
mL) was added within 2 h at Ϫ78 °C. The mixture was allowed
to warm to 23 °C (16 h), and poured into cold (ice/water bath)
hydrochloric acid (25 mL, 2 ). The aqueous layer was separated
and extracted with CH2Cl2 (1 ϫ 60 mL). The combined organic
layers were dried (MgSO4), the solvent was removed under vacuum,
and the crude product chromatographed on SiO2 (PE/EA, 1:1, Rf ϭ
0.27) to furnish compound 4d as a colorless oil (7.72 g, 35.5 mmol,
Synthesis of α-Acetyllactams
The preparation of α-acetyllactams appears to be consid-
erably simpler than the formation of the corresponding lac-
tones, which we reported recently.[23] The parent lactams 9a,
9b (Scheme 6) are deprotonated with LDA in THF at Ϫ78
°C followed by addition of MeOAc and warming of the
reaction mixture to ambient temperature. The use of AcCl
and Ac2O as acylating reagents results in significantly
lower yields.
1
71%). H NMR (CDCl3, 500 MHz): δ ϭ 2.00Ϫ2.08 (m, 1 H), 2.45
(s, 3 H), 2.52Ϫ2.59 (m, 1 H), 3.21Ϫ3.26 (m, 1 H), 3.32Ϫ3.37 (m,
1 H), 3.67Ϫ3.70 (m, 1 H), 4.43 (d, J ϭ 14.7 Hz, 1 H), 4.51 (d, J ϭ
14.7 Hz, 1 H), 7.23Ϫ7.38 (m, 5 H) ppm. 13C{1H} NMR (CDCl3,
125 MHz): δ ϭ 19.4 (CH2), 30.0 (CH), 44.8 (CH2), 47.0 (CH2),
55.7 (CH3), 127.6 (CH), 127.9 (CH), 128.6 (CH), 135.1 (C), 169.7
(CO), 203.6 (CO) ppm. IR (neat): ν˜ ϭ 2920 (m), 1720 (vs), 1680
(vs), 1490 (s), 1450 (s), 1430 (s), 1350 (s) cmϪ1. HRMS (70 eV, EI):
calcd. 217.1103 for C13H15NO2, found 217.1102 [Mϩ]. C13H15NO2
(217.26): calcd. C 71.87, H 6.96, N 6.45; found C 71.43, H 7.00,
N 6.45.
Scheme 6. Synthesis of α-acetyllactams 4d, 4e by deprotonation of
lactams 9a, 9b followed by treatment with MeOAc
α-Acetyl-N-benzyl-δ-valerolactam (4e): According to the procedure
described for 4d, N-benzylvalerolactam 9b (7.80 g, 41.2 mmol) was
converted into crude 4e by adding LDA (20.6 mL, 41.2 mmol, 2.0
) and MeOAc (6.50 mL, 6.03 g, 82.0 mmol). Purification by chro-
matography (SiO2, PE/EA, 1:1, Rf ϭ 0.21) gave compound 4e as a
Conclusion
Cerium() chloride heptahydrate is a readily available
and nontoxic catalyst for the α-hydroxylation of β-dicar-
bonyl compounds. The oxidant is molecular oxygen and the
solvent 2-propanol. With regard to economical and ecolo-
gical considerations both are ideal. For most cyclic sub-
strates 4aϪ4i, ranging from β-oxo esters, β-oxolactones and
β-oxolactams to β-dicarbonyl compounds, quantitative
conversion combined with complete chemoselectivity is
achieved, resulting in a very simple workup procedure: Fil-
1
yellow oil (6.09 g, 26.4 mmol, 64%). H NMR (CDCl3, 500 MHz),
mixture of keto and enol tautomer (ratio 1:2); Keto Form: δ ϭ
1.67Ϫ1.76 (m, 1 H), 1.83Ϫ1.92 (m, 2 H), 2.08Ϫ2.16 (m, 1 H), 2.37
(s, 3 H), 3.17Ϫ3.26 (m, 2 H), 3.54Ϫ3.60 (m, 1 H), 4.53 (d, J ϭ
14.6 Hz, 1 H), 4.66 (d, J ϭ 14.6 Hz, 1 H), 7.23Ϫ7.36 (m, 5 H) ppm;
Enol Form: δ ϭ 1.76Ϫ1.82 (m, 2 H), 1.95 (s, 3 H), 2.37Ϫ2.42 (m,
2 H), 3.22 (t, J ϭ 5.7 Hz, 2 H), 4.60 (s, 2 H), 7.23Ϫ7.36 (m, 5 H),
14.96 (s, 1 H) ppm. 13C{1H} NMR (CDCl3, 125 MHz) ppm: Keto
tration of the reaction mixture through SiO2 removes all Form: δ ϭ 21.4 (CH2), 24.1 (CH2), 30.6 (CH3), 47.6 (CH2), 50.8
(CH2), 56.0 (CH), 127.9 (CH), 128.4 (CH), 129.1 (CH), 137.2 (C),
167.0 (C), 206.3 (C) ppm; Enol Form: δ ϭ 19.0 (CH3), 23.02 (CH2),
24.6 (CH2), 47.7 (CH2), 50.2 (CH2), 96.1 (C), 127.7 (CH), 128.2
(CH), 129.0 (CH), 137.7 (C), 169.7 (C), 170.9 (C) ppm. IR (neat):
ν˜ ϭ 2920 (m), 2840 (m), 1720 (vs), 1640 (vs) cmϪ1. MS (70 eV, EI):
m/z (%) ϭ 231 (96) [Mϩ], 216 (10), 188 (70), 91 (100). C14H17NO2
(231.29): calcd. C 72.70, H 7.41, N 6.06; found C 72.75, H 7.45,
N 6.04.
metal-containing materials and yields analytically pure
products. Unfortunately, for acyclic substrates 4jϪ4p the
chemoselectivity is significantly lower. Only β-oxo esters
and β-diketones bearing an α-alkyl substituent give prod-
ucts with yields up to 50% after chromatography. For a
number of other substrates either no conversion (4qϪ4u,
8aϪ8e) or the formation of very complex reaction mixtures
(4vϪ4z) is observed. We propose a mechanism for the α-
hydroxylation which is in accordance with the stoichi-
ometry calculated from the oxygen uptake during the reac-
tion.
General Procedure for the Cerium-Catalyzed α-Hydroxylation: The
β-dicarbonyl compound 4 (1 equiv.) was added to a suspension of
CeCl3·7H2O (0.05 equiv.) in iPrOH (ca. 3 ). The flask was evacu-
ated twice to ca. 500 mbar, each time flushed with O2, and the
mixture was then stirred for 16 h while a slow stream of O2 (ca. 50
cm3 hϪ1) was passed through.
Experimental Section
Workup Procedure A: The mixture was filtered through ca. 2 cm of
SiO2 (washing with EA or CH2Cl2). All volatiles were removed
from the filtrate under high vacuum to yield the α-hydroxylated
products 5.
General: Starting materials 4a, 4c, 4gϪ4i, 4q, 4wϪ4y, 8d, and 8e are
commercially available. The following compounds were prepared
according to literature procedures: 4b,[24] 4f, 4v, 8c,[23] 4j, 4k,[25]
4l,[26] 4m,[27] 4n,[28] 4o,[29] 4p,[30] 4r,[31] 4s,[32] 4t,[33] 4u,[34] 4z,[35] 8a,[36]
8b,[37] 9a,[38] and 9b.[39] Column chromatography was carried out
using Merck SiO2 60 with hexanes (PE, b.p. 40Ϫ60 °C), and EtOAc
(EA) as eluents. LDA (2.0  in THF/heptane/ethylbenzene) was
purchased from Aldrich Chemical Co. 13C NMR multiplicities were
determined with DEPT experiments.
Workup Procedure B: The solvents were evaporated and the residue
submitted to column chromatography (SiO2, PE/EA) to yield the
products 5.
Workup Procedure C: The solvents were evaporated and the residue
submitted to Kugelrohr distillation under vacuum to yield the
products 5.
α-Acetyl-N-benzyl-γ-butyrolactam (4d): A solution of pyrrolidone
9a (8.77 g, 50.0 mmol) in anhydrous THF (30 mL) was added at
Ϫ78 °C within 4 h (syringe pump) to a solution of LDA
Ethyl 1-Hydroxy-2-oxo-1-cyclopentanecarboxylate (5a): According
to the General Procedure, 4a (260 mg, 1.66 mmol) and
428
Eur. J. Org. Chem. 2003, 425Ϫ431

Preparation of Acyloins from β-Dicarbonyl Compounds
FULL PAPER
CeCl3·7 H2O (30 mg, 0.081 mmol) in iPrOH (0.5 mL) were con-
verted to yield 5a after Workup A as a colorless liquid (282 mg,
1.64 mmol, 99%). All data are in accordance with the literature.[17]
1.27 mmol, 78%). All data are in accordance with the literature.[12]
Ethyl 1-Hydroxy-2-oxo-1-cyclohexanecarboxylate (5h): According
to the General Procedure, 4h (1.089 g, 6.380 mmol) and
CeCl3·7 H2O (119 mg, 0.320 mmol) in iPrOH (2.0 mL) were con-
verted to yield 5h after Workup C (110Ϫ180 °C/2 mbar) as a color-
less liquid (696 mg, 3.74 mmol, 59%). All data are in accordance
with the literature.[17]
L-Menthyl 1-Hydroxy-2-oxo-1-cyclopentanecarboxylate (5b): Ac-
cording to the General Procedure, 4b (202 mg, 0.758 mmol) and
CeCl3·7 H2O (15 mg, 0.040 mmol) in iPrOH (0.5 mL) were con-
verted to yield 5b after Workup A as a colorless liquid (211 mg,
0.747 mmol, 99%). All data are in accordance with the literature.[19]
2-Acetyl-2-hydroxycyclohexanone (5i): According to the General
Procedure, 4i (1.051 g, 7.498 mmol) and CeCl3·7 H2O (140 mg,
0.376 mmol) in iPrOH (2.4 mL) were converted to yield 5i after
Workup C (60Ϫ160 °C/2 mbar) as a colorless liquid (592 mg,
3.79 mmol, 51%). All data are in accordance with the literature.[9d]
Methyl 1-Hydroxy-2-oxo-1-cycloheptanecarboxylate (5c): Accord-
ing to the General Procedure, 4c (274 mg, 1.61 mmol) and
CeCl3·7 H2O (30 mg, 0.081 mmol) in iPrOH (0.5 mL) were con-
verted to yield 5c after Workup A as a colorless liquid (287 mg,
1.54 mmol, 96%). All data are in accordance with the literature.[17]
Ethyl 2-Hydroxy-2-methyl-3-oxobutanoate (5j): According to the
General Procedure, 4j (235 mg, 1.62 mmol) and CeCl3·7 H2O
(30 mg, 0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5j
after Workup B (PE/EA, 5:1, Rf ϭ 0.19) as a colorless liquid
(75 mg, 0.47 mmol, 29%). All data are in accordance with the liter-
ature.[15]
α-Acetyl-N-benzyl-α-hydroxy-γ-butyrolactam (5d): According to the
General Procedure, 4d (328 mg, 1.51 mmol) and CeCl3·7 H2O
(30 mg, 0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5d
after Workup A as a colorless crystalline solid (255 mg, 1.09 mmol,
1
72%). M.p. 84.5 °C. H NMR (CDCl3, 500 MHz): δ ϭ 2.10 (ddd,
J ϭ 13.5, J ϭ 8.8, J ϭ 7.8 Hz, 1 H), 2.30 (s, 3 H), 2.44 (ddd, J ϭ
13.5, J ϭ 6.9, J ϭ 3.7 Hz, 1 H), 3.27Ϫ3.35 (m, 2 H), 4.44 (d, J ϭ
14.7 Hz, 1 H), 4.57 (d, J ϭ 14.7 Hz, 1 H), 5.07 (br. s, 1 H),
7.19Ϫ7.36 (m, 5 H) ppm. 13C{1H} NMR (CDCl3, 125 MHz): δ ϭ
25.47 (CH3), 31.27 (CH2), 43.87 (CH2), 47.77 (CH2), 84.03 (C),
128.29 (CH), 128.47 (CH), 129.23 (CH), 135.67 (CH), 172.21 (C),
Ethyl 2-Ethyl-2-hydroxy-3-oxobutanoate (5k): According to the
General Procedure, 4k (255 mg, 1.61 mmol) and CeCl3·7 H2O
(30 mg, 0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5k
after Workup B (PE/EA, 10:1, Rf ϭ 0.13) as a colorless liquid
(89 mg, 0.51 mmol, 32%). All data are in accordance with the liter-
ature.[4b]
208.21 (C) ppm. IR (KBr): ν˜ ϭ 3260 (br, s), 1714 (s), 1676 (s) cmϪ1
.
MS (70 eV, EI): m/z (%) ϭ 233 (5) [Mϩ], 217 (67), 191 (17), 174
(90), 91 (100). C13H15NO3 (233.26): calcd. C 66.94, H 6.48, N 6.00;
found C 67.20, H 6.53, N 5.90.
Ethyl 2-Acetyl-2-hydroxyhexanoate (5l): According to the General
Procedure, 4l (302 mg, 1.62 mmol) and CeCl3·7 H2O (30 mg,
0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5l after
Workup B (PE/EA, 10:1, Rf ϭ 0.14) as a colorless liquid (149 mg,
1
α-Acetyl-N-benzyl-α-hydroxy-δ-valerolactam (5e): According to the
General Procedure, 4e (358 mg, 1.55 mmol) and CeCl3·7 H2O
(29 mg, 0.078 mmol) in iPrOH (0.5 mL) were converted to yield 5e
after Workup A as a colorless crystalline solid (327 mg, 1.32 mmol,
0.737 mmol, 45%). H NMR (CDCl3, 300 MHz): δ ϭ 0.90 (t, J ϭ
7.1 Hz, 3 H), 1.30 (t, J ϭ 7.1 Hz, 3 H), 1.16Ϫ1.40 (m, 4 H),
1.85Ϫ1.96 (m, 1 H), 2.04Ϫ2.23 (m, 1 H), 2.29 (s, 3 H), 4.17 (s, 1 H),
4.21Ϫ4.36 (m, 2 H) ppm. 13C{1H} NMR (CDCl3, 75 MHz): δ ϭ
13.87 (CH3), 14.06 (CH3), 22.69 (CH2), 24.68 (CH3), 25.22 (CH2),
34.94 (CH2), 62.58 (CH2), 84.24 (C), 171.02 (C), 205.11 (C) ppm.
IR (neat): ν˜ ϭ 3493 (br, s), 2961 (s), 2933 (s), 1749 (s), 1719 (vs)
cmϪ1. MS (70 eV, EI): m/z (%) ϭ 202 (1) [Mϩ], 160 (100), 85 (64),
43 (27). C10H18O4 (202.24): calcd. C 59.39, H 8.97; found C 59.30,
H 8.99.
1
85%). M.p. 102 °C. H NMR (CDCl3, 300 MHz): δ ϭ 1.83Ϫ1.99
(m, 3 H), 2.17Ϫ2.28 (m, 1 H), 2.30 (s, 3 H), 3.21Ϫ3.35 (m, 2 H),
4.50 (d, J ϭ 14.7 Hz, 1 H), 4.59 (s, 1 H), 4.75 (d, J ϭ 14.7 Hz, 1 H),
7.26Ϫ7.38 (m, 5 H) ppm. 13C{1H} NMR (CDCl3, 75 MHz): δ ϭ
18.62 (CH2), 24.59 (CH3), 30.67 (CH2), 47.06 (CH2), 50.71 (CH2),
79.40 (C), 127.68 (CH), 127.96 (CH), 128.77 (CH), 136.19 (C),
169.13 (C), 208.12 (C) ppm. IR (KBr): ν˜ ϭ 3316 (br, s), 2940 (m),
2875 (m), 1716 (s), 1622 (vs) cmϪ1. MS (70 eV, EI): m/z (%) ϭ 247
(2) [Mϩ], 230 (17), 204 (99), 91 (100). C14H17NO3 (247.29): calcd.
C 68.00, H 6.93, N 5.66; found C 67.91, H 6.95, N 5.48.
Ethyl 2-Hydroxy-2-isopropyl-3-oxobutanoate (5m): According to
the General Procedure, 4m (274 mg, 1.59 mmol) and CeCl3·7 H2O
(30 mg, 0.081 mmol) in iPrOH (0.5 mL) were converted to yield
5m after Workup B (PE/EA, 10:1, Rf ϭ 0.16) as a colorless liquid
(40 mg, 0.21 mmol, 13%). In a first fraction (Rf ϭ 0.21) starting
material 4m (174 mg, 1.01 mmol, 63%) was recovered. 1H NMR
(CDCl3, 500 MHz): δ ϭ 0.80 (d, J ϭ 6.7 Hz, 3 H), 0.94 (d, J ϭ
6.7 Hz, 3 H), 1.31 (t, J ϭ 7.2 Hz, 3 H), 2.32 (s, 3 H), 2.72 (sept,
J ϭ 6.8 Hz, 1 H), 4.10 (s, 1 H), 4.22Ϫ4.33 (m, 2 H) ppm. 13C{1H}
NMR (CDCl3, 125 MHz): δ ϭ 14.09 (CH3), 15.97 (CH3), 16.80
(CH3), 25.18 (CH), 33.77 (CH3), 62.59 (CH2), 87.79 (C), 170.84
(C), 205.83 (C) ppm. IR (ATR): ν˜ ϭ 3482 (br, s), 2972 (s), 1721
(vs), 1261 (s), 1177 (s), 1155 (s) cmϪ1. MS (70 eV, EI): m/z (%) ϭ
189 (10) [MHϩ], 171 (13), 146 (34), 115 (72), 71 (100), 43 (90).
C9H16O4 (188.22): calcd. C 57.43, H 8.57; found C 57.63, H 8.52.
α-Acetyl-α-hydroxy-ε-caprolactone (5f): According to the General
Procedure, 4f (250 mg, 1.60 mmol) and CeCl3·7 H2O (30 mg,
0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5f after
Workup A as a colorless crystalline solid (227 mg, 1.32 mmol,
83%). M.p. 70Ϫ72 °C. 1H NMR (CDCl3, 500 MHz): δ ϭ
1.74Ϫ1.88 (m, 3 H), 1.94Ϫ2.06 (m, 2 H), 2.09Ϫ2.17 (m, 1 H), 2.39
(s, 3 H), 4.27 (ddd, J ϭ 12.3, J ϭ 6.9, J ϭ 2.0 Hz, 1 H), 4.60 (s,
1 H), 4.79 (ddd, J ϭ 12.2, J ϭ 9.4, J ϭ 1.4 Hz, 1 H) ppm. 13C{1H}
NMR (CDCl3, 125 MHz): δ ϭ 22.81 (CH2), 24.98 (CH3), 29.07
(CH2), 32.28 (CH2), 69.94 (CH2), 82.35 (C), 172.43 (C), 204.94 (C)
ppm. IR (KBr): ν˜ ϭ 3443 (br, vs), 2974 (m), 2943 (s), 1703 (br, vs)
cmϪ1. MS (CH4, CI): m/z (%) ϭ 173 (21) [MHϩ], 155 (15), 130
(49), 127 (100). C8H12O4 (172.18): calcd. C 55.81, H 7.02; found C
55.81, H 6.98.
Ethyl 2-Hydroxy-2-methyl-3-oxo-3-phenylpropanoate (5n): Accord-
ing to the General Procedure, 4n (270 mg, 1.31 mmol) and
CeCl3·7 H2O (20 mg, 0.054 mmol) in iPrOH (0.5 mL) were con-
verted to yield 5n after Workup B (toluene/EA, 30:1, Rf ϭ 0.16) as
a colorless liquid (87 mg, 0.39 mmol, 30%). In a first fraction (Rf ϭ
0.29) starting material 4n (43 mg, 0.21 mmol, 16%) was recovered.
α-Acetyl-α-hydroxy-γ-butyrolactone (5g): According to the General
Procedure, 4g (208 mg, 1.62 mmol) and CeCl3·7 H2O (30 mg,
0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5g after
Workup B (PE/EA, 1:1, Rf ϭ 0.26) as a yellow oil (183 mg, All data are in accordance with the literature.[19]
Eur. J. Org. Chem. 2003, 425Ϫ431
429

J. Christoffers, T. Werner, S. Unger, W. Frey
FULL PAPER
Table 1. Crystal data and structure refinement for compounds 5dϪ5f
5d
5e
5f
Empirical formula
M [g molϪ1
C13H15NO3
233.26
C14H17NO3
247.29
0.7 ϫ 0.35 ϫ 0.06
monoclinic
C2/c
19.1529(9)
6.6057(3)
23.0387(9)
90
112.938(4)
90
2684.3(2)
8
1.224
0.702
1056
C8H12O4
172.18
0.7 ϫ 0.7 ϫ 0.25
monoclinic
P21/c
8.5823(9)
10.2091(6)
9.9568(9)
90
102.650(7)
90
851.21(13)
4
1.344
0.913
368
]
Crystal size [mm]
Crystal system
Space group
0.3 ϫ 0.15 ϫ 0.10
triclinic
¯
P1
˚
a [A]
6.8367(5)
9.4094(7)
10.7287(7)
68.547(6)
80.050(5)
77.111(6)
623.00(8)
2
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
Z
ρcalcd. [g cmϪ3
]
1.243
0.727
248
µ [mmϪ1
F(000)
]
2Θ range [°]
4.45Ϫ65.00
2027
1395
4.17Ϫ67.98
2289
1335
5.28Ϫ67.98
1441
1300
No. of unique data
No. of data obsd. [I Ͼ 2σ(I)]
No. of parameters refined
R1, wR2 (all data)
R1 (obsd.)
159
168
111
0.0762, 0.2000
0.0593
1.118
0.1198, 0.3304
0.0956
1.161
0.0594, 0.1665
0.0568
1.095
GoF (F2)
Ϫ3
˚
Max./min. electron density [e·A
]
0.207/Ϫ0.221
0.473/Ϫ0.289
0.242/Ϫ0.235
[2]
[3]
D. H. G. Crout, E. R. Lee, D. P. J. Pearson, J. Chem. Soc.,
Chem. Commun. 1990, 331Ϫ333.
Ethyl 2-Benzyl-2-hydroxy-3-oxobutanoate (5o): According to the
General Procedure, 4o (351 mg, 1.59 mmol) and CeCl3·7 H2O
(30 mg, 0.081 mmol) in iPrOH (0.5 mL) were converted to yield 5o
after Workup B (PE/EA, 10:1, Rf ϭ 0.18) as a colorless liquid
(112 mg, 0.47 mmol, 30%). In a first fraction (Rf ϭ 0.38) by-prod-
uct 7b was recovered as a colorless liquid (55 mg, 0.22 mmol, 14%)
and in a second fraction (Rf ϭ 0.19) starting material 4o (147 mg,
0.667 mmol, 42%). All data are in accordance with the literature.[19]
G. Olack, H. Morrison, J. Org. Chem. 1991, 56, 4969Ϫ4971.
[4] [4a]
D. H. G. Crout, D. L. Rathbone, J. Chem. Soc., Chem.
Commun. 1987, 290Ϫ291. [4b] D. H. G. Crout, D. L. Rathbone,
[4c]
Synthesis 1989, 40Ϫ42.
K. Tamaki, J. B. Shotwell, R. D.
White, I. Drutu, D. T. Petsch, T. V. Nheu, H. He, Y. Hirokawa,
H. Maruta, J. L. Wood, Org. Lett. 2001, 3, 1689Ϫ1692.
A. S. Demir, A. Jeganathan, Synthesis 1992, 235Ϫ247.
E. Vedejs, D. A. Engler, J. E. Telschow, J. Org. Chem. 1978,
43, 188Ϫ196.
[5]
[6]
3-Benzyl-3-hydroxy-2,4-pentanedione (5p): According to the Gen-
eral Procedure, 4p (126 mg, 0.662 mmol) and CeCl3·7 H2O (15 mg,
0.040 mmol) in iPrOH (0.35 mL) were converted to yield 5p after
Workup B (PE/EA, 5:1, Rf ϭ 0.34) as a colorless liquid (60 mg,
0.29 mmol, 44%). In a first fraction (Rf ϭ 0.47) by-product 7c was
separated as a colorless liquid (11 mg, 0.049 mmol, 7%). All data
are in accordance with the literature.[19]
[7]
[8]
F. A. Davis, A. C. Sheppard, Tetrahedron 1989, 45, 5703Ϫ5742.
A. J. Hubert, P. S. Starcher, J. Chem. Soc. C 1968, 2500Ϫ2502.
[9] [9a]
G. M. Rubottom, M. A. Vazquez, D. R. Pellegrina, Tetra-
hedron Lett. 1974, 4319Ϫ4322. [9b] A. Hassner, R. H. Reuss, H.
[9c]
W. Pinnick, J. Org. Chem. 1975, 40, 3427Ϫ3429.
A. G.
Brook, D. M. Macrae, J. Organomet. Chem. 1974, 77,
[9d]
C19ϪC21.
R. Z. Andriamialisoa, N. Langois, Y. Langois,
Tetrahedron Lett. 1985, 26, 3563Ϫ3566.
W. Adam, L. Hadjiarapoglou, Top. Curr. Chem. 1993, 164,
45Ϫ62.
W. Adam, J. Bialas, L. Hadjiarapoglou, Chem. Ber. 1991, 124,
2377Ϫ2377.
X-ray Structure Determinations: Crystallographic data for com-
pounds 5d, 5e, and 5f are given in Table 1. The intensities were
measured with a Siemens P4 diffractometer (Cu-Kα, λ ϭ 1.54178
[10]
[11]
˚
A) at 293 K. The structures were solved by using direct methods
and refined against F2 for all observed reflections. CCDC-190172
(5d), -190174 (5e), and -190173 (5f) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge at www.ccdc.cam.ac.uk/conts/retrieving.html [or
from the Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, UK; Fax: (internat.) ϩ 44-1223/336-
033; E-mail: deposit@ccdc.cam.ac.uk].
[12]
[13]
W. Adam, F. Prechtl, Chem. Ber. 1991, 124, 2369Ϫ2372.
W. Adam, A. K. Smerz, Tetrahedron 1996, 52, 5799Ϫ5804.
[14] [14a]
M. Eissen, J. O. Metzger, E. Schmidt, U. Schneidewind,
Angew. Chem. 2002, 114, 402Ϫ425; Angew. Chem. Int. Ed.
[14b]
2002, 41, 414Ϫ436.
2002, 4, 1043Ϫ1045.
J. Muldoon, S. N. Brown, Org. Lett.
[14c]
T. Yokota, S. Sakaguchi, Y. Ishii, J.
[14d]
Org. Chem. 2002, 67, 5005Ϫ5008.
McLurgh, P. Plucinski, Ind. Eng. Chem. Res. 2002, 41,
P. A. Alvarez, D.
2153Ϫ2158.
Acknowledgments
We are grateful to the Fonds der Chemischen Industrie and the
Deutsche Forschungsgemeinschaft for generous support. We thank
Dr. A. Baro for her assistance.
[15]
[16]
X. Baucherel, E. Levoirier, J. Uziel, S. Juge, Tetrahedron Lett.
2000, 41, 1385Ϫ1387.
T. Watanabe, T. Ishikawa, Tetrahedron Lett. 1999, 40,
7795Ϫ7798.
[17]
[18]
J. Christoffers, J. Org. Chem. 1999, 64, 7668Ϫ7669.
K. Hirase, T. Iwahama, S. Sakaguchi, Y. Ishii, J. Org. Chem.
2002, 67, 970Ϫ973.
[1]
J. Zhu, A. J. H. Klunder, B. Zwanenburg, Tetrahedron Lett.
[19]
1994, 35, 2787Ϫ2790.
J. Christoffers, T. Werner, Synlett 2002, 119Ϫ121.
430
Eur. J. Org. Chem. 2003, 425Ϫ431

Preparation of Acyloins from β-Dicarbonyl Compounds
FULL PAPER
J. Christoffers, Synth. Commun. 1999, 29, 117Ϫ122.
C. Weygand, H. Forkel, C. Bischoff, Ber. Dtsch. Chem. Ges.
1928, 61, 687Ϫ690.
P. A. Evans, T. A. Brandt, J. E. Robinson, Tetrahedron Lett.
1999, 40, 3105Ϫ3108.
T. Chiba, T. Takahashi, J. Sakaki, C. Kaneko, Chem. Pharm.
Bull. 1985, 33, 3046Ϫ3049.
S. Yamazaki, Y. Yanase, E. Tanigawa, S. Yamabe, H. Tamura,
J. Org. Chem. 1999, 64, 9521Ϫ9528.
W. B. S. van Liemt, W. F. Steggerda, R. Esmeijer, J. Lugten-
burg, Recl. Trav. Chim. Pays-Bas 1994, 113, 153Ϫ161.
E. C. Taylor, G. H. Hawks III, A. McKillop, J. Am. Chem.
Soc. 1968, 90, 2421Ϫ2422.
M. E. Maier, D. Langenbacher, F. Rebien, Liebigs Ann. 1995,
1843Ϫ1848.
[20] [20a]
[30]
[31]
J. Christoffers, A. Mann, Angew. Chem. 2000, 112,
[20b]
2871Ϫ2874; Angew. Chem. Int. Ed. 2000, 39, 2752Ϫ2754.
J. Christoffers, Synlett 2001, 723Ϫ732.
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[21] [21a]
J. E. Brenner, J. Org. Chem. 1961, 26, 22Ϫ27. [21b] C. W. T.
[21c]
Hussey, A. R. Pinder, J. Chem. Soc. 1962, 1517Ϫ1518.
F.
Karrenbrock, H. J. Schäfer, Tetrahedron Lett. 1978,
1521Ϫ1522.
X.-X. Shi, L.-X. Dai, J. Org. Chem. 1993, 58, 4596Ϫ4598.
J. Christoffers, H. Oertling, P. Fischer, W. Frey, Synlett 2002,
957Ϫ961.
J. Christoffers, N. Önal, Eur. J. Org. Chem. 2000, 1633Ϫ1635.
K. Nakamura, T. Miyai, A. Nagar, S. Oka, A. Ohno, Bull.
Chem. Soc. Jpn. 1989, 62, 1179Ϫ1187.
[22]
[23]
[24]
[25]
[26]
W. B. Renfrow, A. Renfrow, J. Am. Chem. Soc. 1946, 68,
1801Ϫ1804.
Y. Sasson, N. Bilman, J. Chem. Soc., Perkin Trans. 2 1989,
2029Ϫ2033.
M. Nakagawa, Y. Torisawa, T. Hosaka, K. Tanabe, F. Tavet,
M. Aikawa, T. Hino, Heterocycles 1993, 35, 1157Ϫ1170.
Received September 12, 2002
[27]
[28]
W. B. Re n f row, J. Am. Chem. Soc. 1944, 66, 144Ϫ146.
W. Wierenga, H. I. Skulnick, D. A. Stringfellow, S. D. Weed,
H. E. Renis, E. E. Eidson, J. Med. Chem. 1980, 23, 237Ϫ239.
V. A. Martin, D. H. Murray, N. E. Pratt, Y. Zhao, K. F. Albiz-
ati, J. Am. Chem. Soc. 1990, 112, 6965Ϫ6978.
[29]
[O02508]
Eur. J. Org. Chem. 2003, 425Ϫ431
431