Catalytic oxidation of alcohols with allyl diethyl phosphate and palladium acetate

Article abstract of DOI:10.1016/S0022-328X(02)01177-4

Allyl diethyl phosphate (ADP) was found to function as a stoichiometric hydrogen acceptor in a catalytic oxidation reaction of alcohols with Pd(OAc)₂. A variety of acyclic primary and secondary alcohols were oxidized in good yields and under mild conditions to the corresponding aldehydes and ketones, in the presence of Na₂CO₃ or K₂CO₃. Simple aliphatic primary alcohols yielded esters, exclusively. Polar ligand solvents (DMF, DMSO) were found to accelerate the reaction. Slow, but high yield reactions were encountered in THF and acetonitrile as solvents. The reactivity of several other allyl systems serving as H-acceptors, and several Pd compounds serving as catalysts, in the above oxidation reaction, was evaluated. It has been experimentally demonstrated (H-NMR) that ADP is capable of generating a π-allyl-Pd complex using a Pd(0) complex. Consequently, a catalytic cycle was proposed for the above oxidation reaction.

Full text of DOI:10.1016/S0022-328X(02)01177-4

Journal of Organometallic Chemistry 650 (2002) 151–156
www.elsevier.com/locate/jorganchem
Catalytic oxidation of alcohols with allyl diethyl phosphate and
palladium acetate
Youval Shvo *, Vered Goldman-Lev
School of Chemistry, Raymond and Be6erly Sackler School of Exact Sciences, Tel A6i6 Uni6ersity, Tel A6i6 69978, Israel
Received 1 November 2001; accepted 22 December 2001
Abstract
Allyl diethyl phosphate (ADP) was found to function as a stoichiometric hydrogen acceptor in a catalytic oxidation reaction
of alcohols with Pd(OAc)2. A variety of acyclic primary and secondary alcohols were oxidized in good yields and under mild
conditions to the corresponding aldehydes and ketones, in the presence of Na2CO3 or K2CO3. Simple aliphatic primary alcohols
yielded esters, exclusively. Polar ligand solvents (DMF, DMSO) were found to accelerate the reaction. Slow, but high yield
reactions were encountered in THF and acetonitrile as solvents. The reactivity of several other allyl systems serving as
H-acceptors, and several Pd compounds serving as catalysts, in the above oxidation reaction, was evaluated. It has been
experimentally demonstrated (H-NMR) that ADP is capable of generating a p-allyl-Pd complex using a Pd(0) complex.
Consequently, a catalytic cycle was proposed for the above oxidation reaction. © 2002 Elsevier Science B.V. All rights reserved.
Keywords: Oxidation; Alcohols; Palladium acetate; Allyl diethyl phosphate
rates and good yields. Most of the above mentioned
reactions proceeded in satisfactory yields.
1. Introduction
Murahashi et al. [9] have recently demonstrated that
Oxidation of alcohols to ketones and aldehydes is a
various allyl diethyl phosphates (ADPs), in the presence
fundamental reaction in organic chemistry. Economic
of CO, alcohol, and Pd(0) as catalyst, undergo a facile
and environmental aspects dictate a catalytic oxidation
alkyloxy carbonylation, giving rise to unsaturated car-
reaction. Indeed, alcohols can be catalytically oxidized
boxylic acid esters (Eq. (1)). The higher yields and rates
with transition metal complexes of Ru [1], Ni [2], Pt [3],
(compared to other allyl esters and ethers) were at-
and Pd [4]. Several oxidation reactions of alcohols are
tributed to the use of ADPs.
known where a hydrogen acceptor was used, in con-
junction with various Pd catalysts. Thus for example,
bromobenzene [5] was reduced to benzene, with con-
comitant oxidation of an alcohol function. With 1,2-
dichloroethane as hydrogen acceptor, the oxidation
reaction generated ethylene and HCl [6]. Carbon tetra-
chloride was reduced to chloroform upon the oxidation
of an alcohol function [7]. Allyl carbonate undergoes an
intramolecular H transfer, generating the corresponding
ketone, CO2 and propene [8a,8b].
(1)
(2)
(2a)
Noteworthy is the recent discovery by Sheldon co-
workers [8c] of green aerobic oxidation of alcohols with
a Pd(II) phenanthroline complex in a water–alcohol
biphasic system. The reaction proceeded at reasonable
Recently we have discovered [10a] that aldehydes, in
the presence of ADP, Pd(OAc)2, and sodium carbonate,
gave rise to a clean ab-dehydrogenation reaction of
aldehydes (Eq. (2)). Under the above conditions also,
cyclic-ketones gave the corresponding ab-unsaturated
ketones. The yields were moderate to good. The dehy-
* Corresponding author. Tel.: +972-3-640-8687; fax: +972-3-640-
9293.
E-mail address: shvoy@post.tau.ac.il (Y. Shvo).
0022-328X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(02)01177-4

152
Y. Sh6o, V. Goldman-Le6 / Journal of Organometallic Chemistry 650 (2002) 151–156
drogenation of aldehydes under such non-oxidative con-
ditions is significant, and useful.
We have found that Pd(OAc)2, catalytically oxidized
primary and secondary alcohols to the corresponding
aldehydes and ketones, in the presence of stoichiometric
amount of ADP. The allyl moiety of ADP was con-
verted to propylene, as determined by brominating the
evolving gas in CCl4. The stoichiometry of the reaction,
exemplified with 2-octanol as a substrate and ADP, is
described below in Eq. (3). Formally ADP is function-
ing as a hydrogen acceptor. The scope of the oxidation
reaction was studied using a variety of alcohol sub-
strates, and the results are presented in Table 1.
Very recently [10b], we have discovered a reaction
between ADPs and acrylonitrile, catalyzed by
Pd(OAc)2, to give a 1,4-diene (Eq. (2a)). Various substi-
tuted ADP compounds could also be used. This inter-
esting intermolecular metallo-ene reaction involves CꢀC
bond formation, and is currently being further
investigated.
It is evident that ADP is a unique substrate in those
catalytic reactions.
(3)
2. Results and discussion
In the present paper we shall report on still another
new reaction of ADP that is also catalyzed by
Pd(OAc)2. Thus, during the above dehydrogenation
study [10a] we have noted that alcohols, in the presence
of ADP, were readily oxidized to carbonyl compounds.
Thus, cyclohexanol was oxidized to cyclohexanone,
which was further transformed to phenol [10a] with
excess ADP; cholesterol was similarly oxidized to
cholestanone, that was further dehydrogenated to
cholesta-1,4,6-triene-3-one in 75% yield [10a]. This ob-
servation, and the mild reaction condition, prompted us
to study in detail the scope and mechanism of this
oxidation reaction of alcohols with ADP, and
Pd(OAc)2 as catalyst, which is the subject of the present
report.
The following conclusions can be drawn from the
data of Table 1:
1. The oxidation reaction of alcohols proceeds in polar
ligand solvents. Best results were obtained in DMF
(Experiments 1–3). The next best solvent is DMSO
(Experiments 4 and 5). THF and acetonitrile are the
best solvents from the reaction-work-up stand point.
The yields with these solvents are good, but the
reaction is slow (Experiments 1 and 2). Apparently
the rate of the reaction was affected by the polarity,
and ligand power of the reaction solvent.
2. In the present reaction, K2CO3 is a better base than
Na2CO3. It is, most probably, due to the better
solubility of the former in DMF. A reaction time
Table 1
Experimental data for oxidation reactions with Pd(OAc)2 and ADP a
Experiment
Substrate
Base
Time (h)
Solvent
Conversion b (%)
Products (yield %) b
1
2
3
4
5
6
7
8
9
2-Octanol
2-Octanol
2-Octanol
2-Octanol
Na2CO3
Na2CO3
Na2CO3
K2CO3
K2CO3
Na2CO3
Na2CO3
Na2CO3
K2CO3
K2CO3
37
26
8
2
2
12
2.5
46
2
THF
84
84
94
97
76
78
86
65
97
2-Octanone (80)
2-Octanone
2-Octanone (85)
2-Octanone (85)
2-Octanone (64)
Acetophenone
Benzaldehyde (60)
4-Methoxybenzaldehyde
4-Methoxybenzaldehyde (88)
4-Chlorobenzaldehyde (32),
4-Chlorobenzyl-4-chlorobenzoate
1-Octene-3-one (30)
Cinnamaldehyde (90)
(9) Menthone
Diisopropyl ketone
Benzophenone (96)
Benzil (54)
Octyl octanoate
Cyclohexanone, 2-cyclohexene-1-one
CH3CN
DMF
DMF
DMSO
THF
DMF
CH3CN
DMF
DMF
2-Octanol
1-Phenylethanol
Benzyl alcohol
4-OMe-benzyl alcohol
4-OMe-benzyl alcohol
4-Cl-benzyl alcohol
10
4
71, 13
11
12
13
14
15
1-Octene-3-ol
Cinnamyl alcohol
(9) Menthol
Diisopropyl carbinol
Diphenylcarbinol
Benzoin
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
3.25
2
4
1.5
5
6.5
3.5
5
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
64
97
92
98
100
60
16
17
1-Octanol
Cyclohexanol
50
57, 32
18 c
a Reaction conditions: substrate 5 mmol, ADP 6.25 mmol, carbonate base 10 mmol, Pd(OAc)2 0.2 mmol, solvent 4 ml. All reactions were carried
out at 86 °C (bath temperature) under nitrogen blanket.
b Conversions and yields were determined quantitatively by an internal standard (anisole) method, using gas chromatography, and are accurate
to 95%.
c Oil bath temperature was kept at 70 °C.

Y. Sh6o, V. Goldman-Le6 / Journal of Organometallic Chemistry 650 (2002) 151–156
153
Table 2
5. Attempts to oxidize ethyl lactate and 2,2,2-trifl-
uoroethanol failed. Electron withdrawing groups
next to the carbinol C atoms must impede the
oxidation reaction.
Experimental data for oxidation reactions of 2-octanol with
Pd(OAc)2 and various allyl systems a
Experiment
Allyl compound Time (h)
Conversion (%)
Finally, N-phenyl hydroxylamine could also be cata-
lytically oxidized with Pd(OAc)2 as catalyst in the pres-
ence of ADP–K2CO3 under the above reaction
conditions. A transient GC peak of nitrosobenzene
could be detected, which under the basic reaction con-
dition was slowly transformed to azoxybenzene. After
24 h at 75°, azoxybenzene was obtained in 32% yield.
It is of interest to study and compare the efficiency of
other allyl systems in the oxidation reaction of alcohols
with Pd(OAc)2 as catalyst. Several such systems were
examined under identical reaction conditions, and the
results are presented in Table 2. It can be seen from
Table 2 that ADP (Experiment 1) gave the best results.
Allyl methyl carbonate (Experiment 4) also exhibits a
good reactivity, while allyl acetate and bromide (Exper-
iments 2 and 3) are by far inferior.
1
2
3
4
ADP
2.0
7
7
97
18
0
Allyl acetate
Allyl bromide
Allyl methyl
carbonate
6
93
a The reactions were carried out under the same conditions de-
scribed in the footnote to Table 1, in DMF–K2CO3.
Table 3
Experimental data for oxidation reactions of 2-octanol with ADP and
various Pd compounds a
Experiment
Pd catalyst
Time (h)
Conversion (%)
1
2
3
4
Pd(OAc)2
Pd(OAc)2–PPh3
Pd–C (10%)
2.0
6
24
6
97
25
75
93
Examination of the catalytic activity of various Pd
compounds in the oxidation of 2-octanol with ADP
gave the following results (Table 3).
Pd2(dba)3·CHCl3
a The reactions were carried out under the same conditions de-
scribed in the footnote to Table 1, in DMF–K2CO3.
From the data of Table 3, it is evident that the
oxidation reaction proceeds with Pd(II) (Experiments 1
and 2) as well as with Pd(0) (Experiments 3 and 4) as
precatalysts. Thus, Pd(II) precatalyst is presumably re-
duced to Pd(0) in the presence of alcohols. Pd(0) is
most probably the initial active catalytic species (vide
infra). PPh3 (Experiment 2) slows down the oxidation
reaction, most probably through much stronger coordi-
nation to the Pd atom, as compared to a weaker solvent
coordination. It is however gratifying that Pd–C, al-
though slow, is reactive. A heterogeneous catalytic re-
action is usually preferred for practical reasons. When
Pd–C was used in the absence of ADP, no reaction
could be detected, thus reflecting on the similarity
between the heterogeneous and homogeneous reactions.
It will be of interest to briefly touch upon the mecha-
nism of the present reaction. First we have addressed
the question whether ADP is capable of forming a
p-allyl Pd complex. For technical reasons it was
difficult to determine its formation under the reaction
factor of 4 was recorded in experiments 3 and 4,
indicating the superiority of K2CO3.
3. Primary aliphatic alcohols, viz 1-octanol, gave octyl
octanoate as a sole product (Experiment 17). It
must have originated via the formation of octanal,
the subsequent formation of dioctyl hemiacetal, fol-
lowed by its fast oxidation to the ester (the transient
octanal was not dehydrogenated due to the fast
formation of dioctyl hemiacetal). However, no ben-
zyl benzoate ester could be detected with the pri-
mary benzyl alcohol (Experiment 7). This is
attributed to the substantial smaller K(Eq) for hemi-
acetal formation of aromatic aldehydes compared to
aliphatic aldehydes [11]. But 4-chlorobenzyl alcohol
(Experiment 10) gave a mixture of both ester and
aldehyde. The intermediate 4-chlorobenzaldehyde
has a more electrophilic carbonyl C atom compared
to benzaldehyde, thus inducing some hemiacetal
formation, and subsequently an ester. 4-Methoxy-
benzyl alcohol (Experiment 9) was selectively oxi-
dized to 4-methoxybenzaldehyde, faster than
4-chlorobenzyl alcohol (Experiment 10), as expected
for an oxidation reaction.
4. The oxidation of cyclohexanol was not selective.
Cyclohexenone, the dehydrogenation product of cy-
clohexanone, was generated as a by-product (Exper-
iment 18). This is in line with our previous finding
that saturated cyclic ketones were dehydrogenated
substantially faster than acyclic ketones [10a], which
were practically inert. In fact this behavior makes
feasible our oxidation reaction of secondary alco-
hols to ketones (acyclic).
1
conditions. The H-NMR spectra of equimolar quanti-
ties of Pd(PPh3)4 and ADP in CDCl3 or in benzene-d6
at room temperature indicated a gradual disappearance
of the allyl system H signals of the ADP. Broad ill-
defined signals appeared at higher magnetic field. We
have therefore resorted to the more stable and rigid
known complex, (bis(phenylimino)acenaphthene)Pd(0).
It was prepared from Pd2(dba)3·CHCl3 and the
(bis(phenylimino)acenaphthene) ligand [12]. The result-
ing Pd(0) complex was then reacted with ADP in
acetone (Eq. (4)), and after work-up, the H-NMR
spectrum was measured (we were unable to crystallize
the product). The chemical shifts (ppm) of the allyl
system H atoms are given below (Eq. (4)).

154
Y. Sh6o, V. Goldman-Le6 / Journal of Organometallic Chemistry 650 (2002) 151–156
ionic than the halogen complexes. However, more im-
portant, the above results indicate that ADP does form
a y-allyl complex with Pd(0).
Granted that a p-allyl-Pd complex is also formed in
the above described oxidation reaction of alcohols in
the presence of a Pd(0) species, a catalytic cycle may
now be constructed, using well known organometallic
chemistry principles (Scheme 1). The step, following
ADP complexation with LnPd(0), must consist of a
phosphate anion exchange with the alcohol substrate,
RCH2OH (Scheme 1). The equilibrium of this exchange
reaction must be shifted to the right due to the fast
neutralization of the relatively strong diethyl phospho-
ric acid produced (Scheme 1). Significantly, no reaction
was found to take place in the absence of base.
The product of such an exchange reaction is de-
scribed by structures 5a and 5b, two extreme formula-
tions of the alcoholate p-allyl-Pd complex. In the
present experiment, the covalent structure 5b is proba-
bly more important than the ionic structure 5a, on the
basis of the following grounds:
(4)
The resonance signals of the four terminal p-allyl
protons in the product (3.62 ppm) were shifted to a
higher magnetic field relative to the three vinyl protons
of ADP (5.33 ppm), in line with a p-allyl Pd complex
structure. Further support for the above p-allyl struc-
ture was gained by comparing the above chemical shifts
of the product with those of the corresponding known
[13] chloride and bromide complexes, l=3.34 (br, 4H),
5.51 (qn, 1H). The allyl protons of the phosphate
complex resonate at somewhat lower field than those of
the above halogen complexes, implying that the phos-
phate complex is more positively charged, i.e. more
(a) The presence of the strong alkyloxy anion in the
ionic formulation 5a is expected to promote a
nucleophilic attack on the positively charged p-allyl
carbon system that would have eventually yielded
an allyl ether. The latter is not a product of the
present reaction.
(b) Mechanistically, b-elimination should take place in
order for the oxidation reaction to occur. This in
turn requires a structure RCH2OꢀPdꢀ, such as in
the covalent formulation 5b, therefore better de-
scribing the catalytic reactive species.
Thus, b-elimination from 5b generated the carbonyl
product, and an allyl Pd hydride species (Scheme 1).
The latter undergoes now reductive elimination of pro-
pylene (identified) and regeneration of Pd(0), which
resumes the catalytic cycle by recomplexation with
ADP. In the scheme it is assumed that the solvent
ligand L stabilizes the Pd atom in the various
complexes.
Scheme 1. Catalytic cycle for the oxidation of alcohols.
Brief kinetic studies indicated that the oxidation rate
of 2-octanol depends on the concentrations of ADP,
Pd(OAc)2 and the alcohol (Figs. 1–3). This implies that
the formation of the p-allyl-Pd complex (5a–5b) must
be the slow step of the catalytic cycle. The subsequent
b- and reductive eliminations are most probably fast
reaction steps [14].
In conclusion, alcohols may be readily oxidized un-
der mild conditions and satisfactory yields to ketones
or aldehydes with ADP, and Pd(OAc)2 as a precatalyst,
in the presence of a carbonate base. ADP may be
readily prepared from the commercially available di-
ethyl chlorophosphate and allyl alcohol [9]. ADP is a
unique stoichiometric H-acceptor ligand for the above
oxidation reaction. A variety of polar solvents may be
used, and the rate of the reaction is solvent dependent.
Fig. 1. Oxidation rate of 2-octonal as a function of ADP concentra-
tion.

Y. Sh6o, V. Goldman-Le6 / Journal of Organometallic Chemistry 650 (2002) 151–156
155
ADP was prepared from the commercially available
diethyl chlorophosphate and allyl alcohol [9].
All oxidation reactions were carried out under a
nitrogen blanket, and GC monitored their progress. At
the end of the reaction, the catalyst was filtered, the
solvent was removed by evaporation under reduced
pressure, and the residue was analyzed by GC, as well
as by NMR and IR measurements. The oxidation
products are known commercially available com-
pounds, and were identified by GC co-injection with
authentic commercial samples.
3.1.1. Azoxybenzene
N-Phenyl hydroxylamine was prepared according to
the procedure of Vogel [15]. A solution of N-phenyl
hydroxylamine (0.545 g, 5 mol), ADP (1.212 g, 6.25
mol), K2CO3 (1.38 g, 10 mol) and Pd(OAC)2 (0.045 g,
0.2 mol) in DMF (4 ml), was heated overnight at
75 °C. GC injection of the reaction solution after 4 h
indicated a nitrosobenzene signal (co-injection of an
authentic sample) that disappeared at the end of the
reaction to give an azoxybenzene signal, identified by
co-injection of an authentic sample. The solvent was
evaporated in vacuum, and the product was purified by
column chromatography (silica) from pet ether (0.158
g, 32%). MS m/z: [M+] (198), and GC co-injection with
an authentic sample of azoxybenzene.
Fig. 2. Oxidation rate of 2-octonal as a function of octonal concen-
tration.
3.1.2. Octyl octanoate (Experiment 17)
1H-NMR (CDCl3): 0.85–1.61 (28H), 2.29 (t, 7.4 Hz,
2H), 4.06 (t, 6.6 Hz, 2H) ppm. MS m/z: [M++H]
(257), [M+−O(CH2)7CH3] (127).
3.1.3. Bis(phenylimino)acenaphthene)y-allyl palladium
diethylphosphate
Bis(phenylimino)acenaphthene) ligand was prepared
according to
Bis(phenylimino)acenaphthene)
a
literature procedure [12,13].
(0.8 mmol),
Fig. 3. Oxidation rate of 2-octonal as a function of Pd(OAc)2
concentration.
Pd2(dba)3·CHCl3 (0.4 mmol) and ADP (1.12 mmol)
were stirred in acetone (18 ml) for 1 h at room temper-
ature. The solvent was evaporated in vacuum, the
residue washed with ether (3×7 ml), CH2Cl2 (7 ml)
was added to the residue, and the solution was filtered
through celite. Evaporation of the filtrate gave a semi-
In most cases the yields are good. The reaction was
found to proceed via p-allyl-Pd diethyl phosphate inter-
mediate complex. The above oxidation reaction may be
particularly useful in cases where classical oxidative
reaction conditions are prohibitive.
1
solid red residue, H-NMR (CDCl3): 1.16 (t, 7.1 Hz,
6H), 3.62 (m, 4H), 3.9 (qn, 7.0 Hz, 4H), 5.91 (q, 9.7 Hz,
1H), 7.10 (d, 7.6 Hz, 2H), 7.55 (m, 10H), 8.17 (d, 8.3
Hz, 2H) ppm (proper integration of the signals was
obtained upon addition of D2O).
3. Experimental
3.1. General
3.1.4. Kinetic measurements
All kinetic measurements were carried out in 10 ml
volumetric flasks, in DMF, under nitrogen, with mag-
netic stirring, in an oil bath kept at 7092 °C. In all
cases, the molar ratio for K2CO3–ADP was 10:6.25.
Initial reaction rates were measured (GC) for variable
Data related to the quantities of reactants and sol-
vents, reaction temperature and time, as well as percent
conversions and yields are given in Tables 1–3 and
their accompanying footnotes.

156
Y. Sh6o, V. Goldman-Le6 / Journal of Organometallic Chemistry 650 (2002) 151–156
[6] S. Ait-Mohand, F. Henin, J. Muzart, Tetrahedron Lett. 36
(1995) 2473.
[7] H. Magashima, J. Tsuji, Chem. Lett. (1981) 1171.
[8] (a) J. Tsuji, I. Minami, I. Shimizu, Tetrahedron Lett. 25 (1984)
2791;
concentrations of 2-octanol, ADP and Pd(OAc)2. Plots
of the constant initial reaction rates versus variable
concentrations were made for each of the above vari-
ables, and are presented in Figs. 1–3.
(b) I. Minami, J. Tsuji, Tetrahedron 43 (1987) 3903;
(c) G.J. Brink, I.W.C.E. Arends, R.A. Sheldon, Science 287
(2000) 1636.
References
[9] S.-I. Murahashi, Y. Imada, Y. Taniguchi, S. Higashiura, J. Org.
Chem. 58 (1993) 1538.
[1] S-I. Murahashi, K. Ito, T. Naota, Y. Maeda, Tetrahedron Lett.
22 (1981) 5327.
[2] M.P. Doyle, W.J. Patrie, S.B. Williams, J. Org. Chem. 44 (1979)
2955.
[3] M. Schamma, H.R. Rodrignes, Tetrahedron 24 (1968) 6583.
[4] T.F. Blackburn, J. Schwartz, J. Chem. Soc. Chem. Commun.
(1977) 157.
[5] (a) T. Yoshinao, Y. Youichi, Y. Zen-Ichi, Tetrahedron Lett. 16
(1979) 1401;
[10] (a) Y. Shvo, A. Arisha, J. Org. Chem. 63 (1998) 5640;
(b) Y. Shvo, A. Arisha, J. Org. Chem. 66 (2001) 4921.
[11] N. Menashe, Y. Shvo, Organometallics 10 (1991) 3885.
[12] R. Van Asselt, C.J. Elsevier, W.J. Smeets, A.L. Spek, R.
Benedix, Recl. Trav. Chim Pays-Bas 113 (1994) 88.
[13] R. Van Asselt, K. Vrieze, C.J. Elsevier, J. Organomet. Chem.
480 (1994) 27.
[14] J.P. Collman, et al., Principles and Applications of Organotran-
sition Metal Chemistry, 2nd Ed., University Science Books,
1987, p. 323.
(b) T. Yoshinao, Y. Yoshimi, I. Kenji, Y. Youichi, Y. Zen-ichi,
J. Am. Chem. Soc. 48 (1983) 1286;
[15] A. Vogel, Text Book of Practical Organic Chemistry, 4th Ed.,
Longman Group Ltd., 1978, p. 722.
(c) T. Yoshinao, I. Kenji, Y. Yoshimi, Y. Zen-ichi, Tetrahedron
Lett. 22 (1981) 1801.