A new route to polykis(dialkylamino)benzenes and -naphthalenes based on protodefluorination of electron-rich fluoroaromatics: Anion radicals of arenes as a simple and effective alternative to 'classical' LAH-based systems

Article abstract of DOI:10.1055/s-2005-921755

A simple and effective procedure for protodefluorination of electron-rich fluoroaromatic compounds has been developed. It operates with aromatic anion radicals as reducing agents and shows superior results over 'classical' lithium aluminum hydride based s

Full text of DOI:10.1055/s-2005-921755

PAPER
97
A New Route to Polykis(dialkylamino)benzenes and -naphthalenes Based on
Protodefluorination of Electron-Rich Fluoroaromatics: Anion Radicals of
Arenes as a Simple and Effective Alternative to ‘Classical’ LAH-Based
Systems
A
N
ew Route to
P
l
olykis(d
a
ialkylamin
d
o)benzene
s
a
nd
-
na
m
phthalenes ir Igorevich Sorokin,*a Valery Anatolievich Ozeryanskii,a Gennady Sergeevich Borodkinb
a
Department of Organic Chemistry, Rostov State University, Zorge street 7, 344090 Rostov-on-Don, Russia
Fax +7(863)2974156; E-mail: vsorokin@aaanet.ru
Institute of Physical and Organic Chemistry, Stachki ave. 194/2, 344090 Rostov-on-Don, Russia
b
Received 14 June 2005; revised 14 July 2005
Dedicated to Prof. Alexander F. Pozharskii on the occasion of his birthday
NR2
NR2
NR2
NR2
Abstract: A simple and effective procedure for protodefluorination
of electron-rich fluoroaromatic compounds has been developed. It
operates with aromatic anion radicals as reducing agents and shows
superior results over ‘classical’ lithium aluminum hydride based
systems.
F
F
F
F
F
NR2
F
R2N
Key words: fluorinated amino arenes, aromatic anion radicals, pro-
todefluorination, lithium aluminum hydride, electron transfer
1 NR2 = NMe2
5 NR2 = NMe2
2 NR2 = Piperidin-1-yl
3 NR2 = Pyrrolidin-1-yl
4 NR2 = NBu2
6 NR2 = Piperidin-1-yl
7 NR2 = Pyrrolidin-1-yl
8 NR2 = NBu2
The carbon–fluorine bond is known as one of the stron-
gest among all organoelement bonds. The dissociation en-
ergy is 116 kcal/mol in contrast to 85 kcal/mol for C–C,
99 kcal/mol for C–H and 79 kcal/mol for C–Cl bonds.1
This is the main reason why only a limited range of re-
agents is available for its hydrogenolysis. The most effec-
tive and widely used reagents are systems based on
lithium aluminum hydride (LAH).2 As an alternative for
C–F bond cleavage, Ni(0)/N-heterocyclic carbene com-
plexes, recently suggested by Fort et al., should also be
mentioned.3
Figure 1
F
F
F
R2N
R2N
NR2
NR2
9 NR2 = NMe2
10 NR2 = Piperidin-1-yl
11 NR2 = Pyrrolidin-1-yl
F
Figure 2
Hence, we have developed an effective protodefluorina-
tion procedure should result in a good two-step (nucleo-
philic substitution–hydrogenolysis) alternative to
synthesize polykis(dialkylamino)arenes, easily accessible
from hexafluorobenzene and octafluoronaphthalene com-
pounds 1–11.
Earlier, we developed an effective approach to a wide va-
riety of 1,4-bis(dialkylamino)tetrafluoro- (1–4) and
1,2,4,5-tetrakis(dialkylamino)difluorobenzenes
(5–8)
(Figure 1)4 as well as 2,3,6,7-tetrakis(dialkylamino)tet-
rafluoronaphthalenes (9–11) (Figure 2),5 via nucleophilic
substitution of fluorines in hexafluorobenzene and oc-
tafluoronaphthalene. Continuing our investigations in the
field of amino-substituted perfluoroaromatics, we noticed
that their defluorinated analogues, due to their easily oxi-
dizable nature, find application, for example, in the prep-
aration of different magnetic and conductive materials,6 or
as mediators of glucose oxidase in glucose amperometric
detection.7
We started our efforts using LAH alone, by refluxing
amines 1–11 for 24 hours with ten equivalents of LAH per
fluorine atom in DME (smaller amounts of LAH led to in-
complete defluorination). These condition turned out to be
effective only in the case of dimethylamino naphthalene 9
(Equation 1).
F
F
Me2N
Me2N
NMe2
NMe2
Me2N
Me2N
NMe2
NMe2
LAH
DME, 24 h,
reflux
One of the main synthetic approaches to such compounds
is still a step-by-step functionalization, (e.g., arene nitra-
tion–reduction–alkylation technique).8 Cross-coupling
F
F
72%
9
12
methodology, becoming more and more popular for its Equation 1
high effectiveness and selectivity, is limited by the intro-
duction of primary and cyclic secondary amino groups
The nature of the solvent had a similar effect on this trans-
formation as that observed in other LAH reductions.10
Thus, in dioxane, the reduction 9 → 12 did not occur, and
in diethyl ether, only traces of product 12 formed after 24
hours, whereas in THF, amine 12 was almost the sole
product, and was the only product when DME was used.
into aromatic compounds.9
SYNTHESIS 2006, No. 1, pp 0097–0102
Advanced online publication: 16.12.2005
DOI: 10.1055/s-2005-921755; Art ID: Z11505SS
© Georg Thieme Verlag Stuttgart · New York
0
6
.
0
1
.2
0
0
6

98
V. I. Sorokin et al.
PAPER
Benzene derivatives, as well as tetraaminonaphthalenes Unfortunately, we were unable to isolate any products
10 and 11, remained unchanged under the conditions from the reaction with polyaminonaphthalenes 10 and 11,
shown in Equation 1.
despite the 100% conversion of the starting compounds.
Extraction of the reaction mixture with diethyl ether, even
using standard precautions and argon atmosphere, led to
complete tarring of the extract. Such behavior arises pos-
sibly from a high electron density in the naphthalene ring
of the defluorinated products 23 and 24 (Figure 3). This
results in a much better conjugation of the piperidino and
pyrrolidino groups with the aromatic system [if compared
to 12 (Table 1)], along with lengthy p–systems which fa-
vor the oxidation of amines 23 and 24.
Interestingly, tetraamine 12 was earlier synthesized in a
four-step procedure starting from 2,7-dihydroxynaphtha-
lene in a total yield of about 10%.11 Taking into account
that compound 9 is prepared from octafluoronaphthalene
in 80% yield, our methodology is at least six times more
effective.
Semi-empirical AM1 calculations12 helped to gain deeper
insight into this phenomenon (Table 1). As can be seen,
tetraamine 9, among others, has the highest sum of total
charges on the aromatic carbons (as the measure of elec-
tron deficiency in the aromatic ring) and the lowest
LUMO energy (as the measure of electron acceptor ability
of the molecule). As a result, derivative 9 should be the
most reactive towards reducing species.
NR2
R2N
R2N
NR2
NR2
F
F
NR2
23 NR2 = Piperidin-1-yl
24 NR2 = Pyrrolidin-1-yl
25
It is well-known that the combination of LAH with some
transition-metal salts can significantly increase the reduc-
ing power of such systems.10 The most active additives to-
ward protodefluorination are known to be cerium(III)
chloride13 and, as suggested recently, niobium pentachlo-
ride.14 Nevertheless, up to now, these salts were not tested
in the defluorination of strongly electron-enriched arenes,
nor was their hydrogenolysis strength compared.
Figure 3
In the case of diaminobenzenes 1–4, completion of the re-
action after 12 hours of heating allowed the detection of
p-difluoro-substituted derivatives of type 25 along with
other products, as evidenced by NMR techniques. In par-
ticular, both 1H and 19F spectra of 25 (NR2 = piperidin-1-
yl) exhibit a doublet of doublets near d = 6.6 (1H) and
d = –150.4 (19F) with coupling constants of 9.1 Hz and
In the present work, the protodefluorination with cerium
trichloride as additive has been conducted according to a
literature procedure,13 using 1.5 equivalents of the salt, 10
equivalents of LAH per fluorine atom, and DME as the
solvent. The results obtained after 24 hours of refluxing
are presented in Table 2. Evidently, the addition of the ce-
rium compound significantly increases the reducing activ-
ity of LAH and allows the inclusion of benzenes 1–5 (but
not 6–8) in the transformation.
3
5
15.2 Hz, characteristic for JHF and JFF, respectively.15
These observations helped to shed light on the defluorina-
tion sequence of fluorobenzenes 1–4. Spectral data of the
reaction mixture of naphthalenes 9–11 were too complex
and did not allow the determination of the defluorination
order.
Tetrakis(dimethylamino)benzene 5 gave a mixture of 17
and 18 (Table 2). Compound 17 shows doublets in the 1H
as well as the 19F NMR spectra with the characteristic cou-
pling constant of 5JHF = 1.9 Hz.15 The relative distribution
of the products in the reaction mixture did not change sig-
nificantly even after 48 hours of heating, possibly due to
some loss of LAH in side reactions.
Table 1 Selected AM1-Calculated Parameters of Arenes 1–11
Com- Average twisting of the
Sum of total charges LUMO
pound NR2 groups relative to the on aromatic carbons energy
aromatic ring plane (°)
(eV)
1
2
47
47
39
49
77
79
77
79
93
78
65
0.44
0.38
0.38
0.40
0.56
0.34
0.51
0.38
0.94
0.74
0.68
–0.63
–0.53
–0.51
–0.53
–0.23
–0.17
–0.15
–0.06
–1.10
–0.86
–0.73
Protodefluorination with niobium pentachloride as an ad-
ditive was performed in a manner similar to the literature
procedure14 using 0.1 equivalent of the niobium(V) chlo-
ride together with 10 equivalents of LAH per fluorine
atom (Table 2). In general, this system has a reducing
power similar to that of the cerium-based composition.
Again, tetraamines 6–8 remained unchanged. It should be
stressed, however, that the reaction rate increased signifi-
cantly, which can be seen especially well by comparing
the results on protodefluorination of compound 5
(Table 2). As before, we were unable to isolate any deflu-
orinated products in the case of naphthalenes 10 and 11,
observing only strong tarring. It should also be mentioned
that, as observed for the cerium-based system, the partial-
ly reduced derivatives of type 25 were also detected here.
3
4
5
6
7
8
9
10
11
Synthesis 2006, No. 1, 97–102 © Thieme Stuttgart · New York

PAPER
A New Route to Polykis(dialkylamino)benzenes and -naphthalenes
99
Table 2 Protodefluorination of Polyamines 1–11 under Various Conditions
Substrate
Product(s)
Yields (%)a
CeCl3/LAH/DME
NbCl5/LAH/DME
100 (61)
Na+Ar–/DME
100 (65)b
1
100 (59)
100 (59)
100 (57)
100 (56)
95
Me2N
NMe2
13
2
3
4
5
100 (59)
100 (59)
100 (61)
78c
100 (63)b
100 (60)b
100 (62)b
0b
N
N
14
N
N
15
Bu2N
NBu2
16
F
Me2N
NMe2
NMe2
Me2N
17
5
22c
100 (68)b
100 (77)e
Me2N
NMe2
NMe2
Me2N
18
6
0d
0d
N
N
N
N
19
7
7
0d
0d
100 (56)e
100 (63)f
F
N
N
N
N
20
0d
0d
N
N
N
N
21
8
9
0d
0d
100 (60)e
100 (65)b
Bu2N
NBu2
NBu2
Bu2N
22
100 (61)
100 (63)
Me2N
NMe2
NMe2
Me2N
12
a Isolated yields are given in parentheses.
b Ar = Naphthalene.
c 3 Equiv of NbCl5 were used to compare with CeCl3 experiments.
d Starting compound was recovered.
e Ar = Biphenyl.
f Ar = 4-Methylbiphenyl.
Synthesis 2006, No. 1, 97–102 © Thieme Stuttgart · New York

100
V. I. Sorokin et al.
PAPER
Interestingly, despite the similarity of the two mentioned can be explained based on data from Arnett and co-work-
additives, completely different mechanisms of action ers.18 They showed that the sodium cation is much stron-
were suggested in the literature. For cerium trichloride, ger when solvated by DME than by THF, with solvation
the single-electron transfer (SET) process with low-valent energies of –0.45 kcal/mol and –0.08 kcal/mol, respec-
cerium as the actual reducing agent was proposed.13 In the tively. Hence, the ion pair between Na+ and the arene an-
case of niobium pentachloride, a nucleophilic substitution ion radical in DME is much more separated, leading to
in the h6-arene complexes of type 26 was postulated to oc- more reactive reducing species. This also explains why in
cur.14
dioxane (solvation energy is still lower, –0.06 kcal/mol)
the starting compounds remain in solution even after 24
hours. We suggested the following reaction mechanism,
presented in Scheme 1, using the example of 1,2,4,5-tet-
rakis(dialkylamino)benzenes and THF.
F
(DME)NbCl3
The mechanistic pathway begins with the electron transfer
from the arene anion radical to the polyamine leading to
the anion radical of 27. Its subsequent transformation to
radical 28, which abstracts a hydrogen atom from the sol-
vent and turns into product 29, terminates the reaction.
26
Figure 4
If the SET mechanism takes place in the defluorination
with cerium(III) chloride, utilization of certain radicals
can lead to the same or even better results, since this will
shorten the electron-transfer pathway from an electron do-
nor to a fluoroaromatic compound.
Studied polyamines can be arranged according to their
protodefluorination potential in the following order: 7 < 8
< 6 < 5 < 3 < 2 ~ 4 < (11, 10) < 9. In general, this row fits
with the sequence predicted on the basis of AM1 calcula-
tions (Table 1), with the exception of tetraamines 5–8.
This deviation compared to the data observed experimen-
tally arises possibly from the lack of a strict geometry pre-
diction for such complexes and, probably more important,
strained systems.
We decided to use sodium naphthalenide as electron
source. This compound can be easily prepared by action
of sodium metal on a naphthalene solution in THF at room
temperature.16 We suggested the general procedure con-
sisting of two simple steps to follow: generation of naph-
thalenide (or other arene anion radicals, see below) (3
equiv per fluorine) and its interaction with selected fluo-
roaromatic compounds (r.t., 12 h).
In conclusion, the proposed protodefluorination technique
is a good alternative both to classical and cross-coupling
methodologies. For example, for 1,4-bis(piperidin-1-
yl)benzene (14), which can be obtained by alkylation of p-
phenylenediamine with 1,5-dibromopentane in 70%
yield,8 or by cross-coupling approach from 1,4-dibro-
mobenzene and piperidine using Pd(dba)2/P(o-tolyl)3 as
catalyst in 63% yield,19 our method (starting from
hexafluorobenzene) gave 50% yield. The difference be-
comes already more important in the case of 1,4-bis(pyr-
Indeed, amines 1–5 and 9–11 undergo C–F bond hydro-
genolysis in the above system; for example, tetraamine 5
gave a mixture of 17 and 18 in 33% and 67%, respective-
ly, indicating the system to be more powerful than LAH-
based mixtures (compare with data from Table 2). We
have also found that DME is superior over THF as a sol-
vent, leading to compound 18 exclusively under similar
conditions (Table 2). Interestingly, the use of sodium
alone in DME did not cause defluorination. However, de-
rivatives 6–8 are still inert to sodium naphthalenide. To
involve them in the protodefluorination sequence, much
stronger electron donors like the anion radical of biphe-
nyl,17 are required. With this anion radical, tetraamines 6
and 8 gave the corresponding di(defluorinated) deriva-
tives 19 and 22. Compound 7 undergoes only monodeflu-
orination leading to 20.
rolidin-1-yl)benzene
(15)
and
1,2,4,5-tetrakis-
(dimethylamino)benzene (18). Compound 15 was ob-
tained from p-phenylenediamine in only 29% yield,8 com-
pared to 47% if prepared by our method; compound 18,
synthesized from 1,2,4,5-tetraaminobenzene in 27%
yield,11 shows comparable results yielding 59% by em-
ploying hexafluorobenzene. A remarkable feature of the
suggested system, along with high effectiveness, is its
customizability. Choosing an appropriated arene allows to
remove but one among several fluorine substituents from
a molecule, as for instance, it was the case in the protode-
fluorination of tetraamine 7 (Table 2).
4-Methylbiphenyl, with an ionization potential lower than
that of biphenyl,17 makes it possible to overcome this ob-
stacle. It is necessary to note that compound 21 shows the
same oxidizable nature as do naphthalenes 23 and 24. In
solution, compound 21 readily becomes green and then
turns to a dark-brown color. As a result, we did not man-
LAH and niobium(V) chloride were purchased from Aldrich. Ceri-
um(III) chloride heptahydrate was purchased from Lancaster and
age to collect its NMR data, but in contrast to compounds made anhydrous according to a literature procedure.20 NMR spec-
troscopy was performed on a Varian Unity-300 spectrometer. Melt-
ing points were obtained on a PTP-1 melting point apparatus. Mass
spectrometry was carried out on a VG AutoSpec instrument. Ele-
mental analyses were obtained from the elemental analysis labora-
23 and 24 the structural information was in part derived
from mass spectra and elemental analysis data of the solid.
As demonstrated, the nature of the arene plays a key role
determining the overall strength of the protodefluorinated tory of Institute of Physical and Organic Chemistry (Rostov-on-
Don). Compounds 1–11 were synthesized in conformity with the
system. The influence of solvent in this transformation
literature,4,5 their properties are identical with those described. The
Synthesis 2006, No. 1, 97–102 © Thieme Stuttgart · New York

PAPER
A New Route to Polykis(dialkylamino)benzenes and -naphthalenes
101
F
F
F
R2N
R2N
NR2
NR2
R2N
R2N
NR2
NR2
Na+
+
+
Na
Ar
Ar
F
27
F
F
R2N
NR2
NR2
R2N
R2N
NR2
NR2
Na+
R2N
– NaF
F
H
H
O
H
H
H
O
H
H
H
H
H
R2N
R2N
NR2
NR2
R2N
R2N
NR2
NR2
~H
+
+
H
H
H
H
H
H
F
F
29
28
+
29
Na
Ar
...
Scheme 1 Proposed mechanism for aromatic anion radical mediated defluorination
correctness of the structures assigned for 12–16 and 18 was proven
by comparison with authentic samples prepared as described in the
literature for 13–158 and for 16,21 and using the Staab method for
compounds 12 and 18.11
1,2,4,5-Tetrakis(pyrrolidin-1-yl)-3-fluorobenzene (20)
Colorless crystals; mp 139–140 °C (from MeOH).
1H NMR (300 MHz, CDCl3): d = 1.86 (m, 16 H, CH2CH2), 3.08 [m,
8 H, N(CH2)2], 3.36 [m, 8 H, N(CH2)2], 5.72 (d, J = 1.5 Hz, 1 H, H-
6).
1,2,4,5-Tetrakis(dibutylamino)difluorobenzene (8)
19F NMR (282 MHz, CDCl3): d = –129.8 (d, J = 1.6 Hz, F-3).
MS (EI, 70 eV): m/z (%) = 372 (100) [M+], 344 (14), 302 (19), 149
Colorless crystals; mp 55–56 °C (from MeOH).
1H NMR (300 MHz, CDCl3): d = 0.86 (t, J = 6.0 Hz, 24 H, CH3),
1.20 (m, 16 H, CH2CH2), 3.03 [t, J = 7.5 Hz, 16 H, N(CH2)2].
(23), 70 (17), 57 (20), 43 (31).
19F NMR (282 MHz, CDCl3): d = –133.57 (s).
MS (EI, 70 eV): m/z (%) = 623 (25) [M+], 622 (76), 579 (30), 566
Anal. Calcd for C22H33FN4: C, 70.93; H, 8.93; N, 15.04. Found: C,
71.20; H, 8.94; N, 15.36.
(100), 479 (31), 465 (38), 365 (17), 57 (22), 43 (42).
1,2,4,5-Tetrakis(pyrrolidin-1-yl)benzene (21)
Anal. Calcd for C38H72F2N4: C, 73.26; H, 11.65; N, 8.99. Found: C,
73.28; H, 11.61; N, 8.89.
Due to the fast oxidation on air and in solution, physical data for the
crude product are given. Light-yellow powder; mp 193–195 °C.
MS (EI, 70 eV): m/z (%) = 354 (100) [M+], 326 (15), 297 (13), 284
(20), 255 (13), 70 (12), 58 (13), 55 (15), 43 (35).
Hydrogenolysis of Fluorinated Amines under the Action of
Aromatic Anion Radicals in DME; General Procedure
Anal. Calcd for C22H34N4: C, 74.53; H, 9.67; N, 15.80. Found: C,
74.40; H, 9.70; N, 15.89.
Finely crashed sodium (3 equiv per fluorine atom) was added to a
solution of the corresponding arene (3 equiv per fluorine atom) in
DME (8 mL) under argon atmosphere. After 2 h of stirring at r.t.,
the solid fluorinated amine (0.1 mmol) was added to the deep green
solution and stirring was continued for additional 12 h. The reaction
solution was then poured on 20% aq HCl (15 mL, Caution, violent
heating!). Some more aq HCl was added to adjust the acidity to pH
2. The precipitated arene was separated by filtration and then, to re-
move arene traces, the mixture was extracted with hexanes (2 × 10
mL). The resulting clear solution was made basic (pH 10–11) using
ammonia and the resulting amines were taken up into Et2O (5 × 4
mL). The ethereal solution was dried (Na2SO4) and quickly evapo-
rated to prevent oxidation of the resulting products.
1,2,4,5-Tetrakis(dibutylamino)benzene (22)
Colorless oil.
1H NMR (300 MHz, CDCl3): d = 0.83 (t, J = 8.3 Hz, 24 H, CH3),
1.23 (m, 16 H, CH2), 1.36 (m, 16 H, CH2), 3.02 [t, J = 8.7 Hz, 16 H,
N(CH2)2], 6.49 (s, 2 H, H-3,6).
Anal. Calcd for C38H74N4: C, 77.75; H, 12.71; N, 9.54. Found: C,
77.71; H, 12.53; N, 9.55.
References
1,2,4,5-Tetrakis(piperidin-1-yl)benzene (19)
Colorless powder; mp 256–257 °C (from MeOH).
1H NMR (300 MHz, CDCl3): d = 1.64 (m, 24 H, CH2CH2CH2), 3.02
[m, 16 H, N(CH2)2], 6.50 (s, 2 H, H-3,6).
MS (EI, 70 eV): m/z (%) = 410 (100) [M+], 205 (14), 84 (17), 55
(17).
(1) Smith, M. B.; March, J. March’s Advanced Organic
Chemistry; John Wiley & Sons: New York, 2001, 22–24.
(2) Krespan, C. G. In Chemistry of Organic Fluorine
Compounds II: A Critical Review; Hudlicky, M.; Pavlath, A.
E., Eds.; American Chemical Society: Washington, 1995,
297–300.
(3) Kuhl, S.; Schneider, R.; Fort, Y. Adv. Synth. Catal. 2003,
345, 341.
Anal. Calcd for C26H42N4: C, 76.05; H, 10.31; N, 13.64. Found: C,
76.25; H, 10.01; N, 13.82.
Synthesis 2006, No. 1, 97–102 © Thieme Stuttgart · New York

102
V. I. Sorokin et al.
PAPER
(13) Imamoto, T.; Takeyama, T.; Kusumoto, T. Chem. Lett.
(4) Sorokin, V. I.; Ozeryanskii, V. A.; Borodkin, G. S. J.
Fluorine Chem. 2005, in press.
1985, 1491.
(5) Sorokin, V. I.; Ozeryanskii, V. A.; Pozharskii, A. F. Eur. J.
Org. Chem. 2004, 766.
(6) (a) Goodson, F. E.; Hauck, S. I.; Hartwig, J. F. J. Am. Chem.
Soc. 1999, 121, 7527. (b) Ito, A.; Ono, Y.; Tanaka, K. J.
Org. Chem. 1999, 64, 8236.
(14) Fuchibe, K.; Akiyama, T. Synlett 2004, 1282.
(15) Everett, S. In Chemistry of Organic Fluorine Compounds II:
A Critical Review; Hudlicky, M.; Pavlath, A. E., Eds.;
American Chemical Society: Washington, 1995, 1037–
1086.
(7) Aleksandrovskii, Y. A.; Bezhikina, L. V.; Rodionov, Y. V.
Biokhimia (Moscow) 1981, 46, 708.
(8) Christensen, J. B.; Schiødt, N.-C.; Bechgaard, K.; Buch-
Rasmussen, T. Acta Chem. Scand. 1996, 50, 1013.
(9) (a) Hartwig, J. F. Angew. Chem. Int. Ed. 1998, 37, 2046.
(b) Prim, D.; Campagne, J.-M.; Joseph, D.; Andrioletti, B.
Tetrahedron 2002, 58, 2041.
(10) Seyden-Penne, J. Reduction by the Alumino- and
Borohydrides in Organic Synthesis; Wiley-VCH: New
York, 1997, 224.
(16) Wang, H. C.; Levin, G.; Szwarc, M. J. Am. Chem. Soc. 1978,
100, 3969.
(17) Rembaum, A.; Eisenberg, A.; Haack, R.; Landel, R. F. J.
Am. Chem. Soc. 1967, 89, 1062.
(18) Arnett, E. M.; Ko, H. C.; Chao, C. C. J. Am. Chem. Soc.
1972, 94, 4776.
(19) Beletskaya, I. P.; Bessmertnykh, A. G.; Guilard, R.
Tetrahedron Lett. 1999, 40, 6393.
(20) Takeda, N.; Imamoto, T. Org. Synth. 1998, 76, 228.
(21) Schröder, U.; Wadhawan, J.; Evans, R. G.; Compton, R. G.;
Wood, B.; Walton, D. J.; France, R. R.; Marken, F.; Page, P.
C. B.; Hayman, C. M. J. Phys. Chem. B 2002, 106, 8697.
(11) Staab, H. A.; Elbl-Weiser, K.; Krieger, C. Eur. J. Org.
Chem. 2000, 327.
(12) Stewart, J. J. P. MOPAC 6.0 Quantum Chemistry Program
Exchange, University of Indiana, Bloomington, Indiana,
USA.
Synthesis 2006, No. 1, 97–102 © Thieme Stuttgart · New York