Gold catalysis: Proof of arene oxides as intermediates in the phenol synthesis

Article abstract of DOI:10.1002/anie.200462672

(Chemical Equation Presented) An elusive intermediate in the gold-catalyzed phenol synthesis was elucidated by modifying the energy profile of this process. Thus, the use of a ligand based on 2-hydroxymethylpyridine enabled the first observation of the in

Full text of DOI:10.1002/anie.200462672

Communications
Reactive Intermediates
II (Scheme 2). According to their calculations, pathway II is
favored, and side products observed in Pt -catalyzed reac-
II
Gold Catalysis: Proof of Arene Oxides as
Intermediates in the Phenol Synthesis**
tions would also correspond to the hydrolysis of intermediate
D of pathway II. However, such side products do not
[
2]
A. Stephen K. Hashmi,*
Matthias Rudolph, Jan P. Weyrauch,
Michael Wꢀlfle, Wolfgang Frey, and
Jan W. Bats
The gold-catalyzed synthesis of highly
substituted arenes 2 or benzofurans 3
from furans 1 has proved to be a
powerful tool for organic synthesis
[
1]
(
Scheme 1). As reported previously,
a number of other transition metals
8
with a d configuration also catalyze
[
1b,2]
this transformation,
but all are sig-
[
1b]
nificantly less active than gold(iii).
We had obtained experimental evi-
dence for an intramolecular migration
of the furan oxygen atom (which
[
1a]
becomes the phenol oxygen atom).
Such a 1,2-transposition suggested the
intermediacy of an arene oxide, but in
our initial publication we only dared to
propose a simple, organic-type mecha-
nism (pathway I via A and B, Sche-
Scheme 2. Possible pathways for the transformation of 1 into 2.
[
1a]
me 2).
Subsequently, Echavarren and co-workers conducted
theoretical studies in which they compared pathways I and
necessarily stem from the catalytic cycle but can be generated
in a competing side reaction. Furthermore, the gold-catalyzed
reactions were highly selective, and such side products were
never observed. Other conceivable pathways would proceed
through alkynyl or vinylidene complexes (I or J, pathway IV),
8
or, if the d precatalyst was reduced in situ, the insertion of a
1
0
2
d
species into the C(sp )ÀO bond (as in the Felkin/Wenkert
[
3]
reaction; pathway III), followed by insertion of the alkyne
to give F.
Herein we describe how the arene oxides/oxepins G/H
can be enriched and detected readily in the reaction mixture
after modification of the energy profile of the reaction by
variation of the ligand in the gold complex. Our efforts to gain
further experimental mechanistic insight formed the starting
point of these studies. Even experiments with substoichio-
1
4
Scheme 1. The gold-catalyzed phenol synthesis. R –R =alkyl, aryl,
5
5
5
6
alkynyl; X=CR , NR , O (three atoms in the linker), ÀCR NR À (four
2
2
atoms in the linker).
[
*] Prof. Dr. A. S. K. Hashmi, Dipl.-Chem. M. Rudolph,
Dipl.-Chem. J. P. Weyrauch, M. Wꢀlfle, Dr. W. Frey
Institut fꢁr Organische Chemie
metric amounts of AuCl (30 mol%) led to no detectable
3
concentration of an intermediate. When a mixture at À208C
Universitꢂt Stuttgart
of AuCl3 (5 mol%) and the substrate was warmed up
1
Pfaffenwaldring 55, 70569 Stuttgart (Germany)
Fax: (+49)711-685-4321
gradually, either no reaction was observed by H NMR
spectroscopy, or the slow formation of 2 at 08C. The breaking
of four bonds and the formation of four new bonds during the
reaction is clearly not a single-transition-state elementary
reaction, and thus the failure to detect any intermediates with
E-mail: hashmi@hashmi.de
Dr. J. W. Bats
Institut fꢁr Organische Chemie und Chemische Biologie
Johann Wolfgang Goethe-Universitꢂt
Marie-Curie-Strasse 11, 60439 Frankfurt am Main (Germany)
AuCl simply meant that the first step was the rate-limiting
3
step. If the reaction proceeded by pathway IV, a primary
kinetic isotope effect should be observable with 1 deuterated
at the alkyne, but this effect was not observed.
[
**] This work was supported by the Fonds der Chemischen Industrie, in
particular through a Chemiefonds-Doktoranden-Stipendium to
J.P.W., and by the AURICAT EU-RTN (HPRN-CT-2002-00174).
To detect intermediates, it is necessary to change the
energy profile of the whole reaction; a later step must have
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2
798
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200462672
Angew. Chem. Int. Ed. 2005, 44, 2798 –2801

Angewandte
Chemie
1
the highest energy of activation. When we used 1a (R = Me,
2
4
R –R = H, X = NTs; Ts = p-toluenesulfonyl) and the com-
[
4]
[5]
plex 4 (X-ray crystal structure shown in Figure 1), we
indeed observed an additional species 5a/6a (Figure 2).
Figure 1. Crystal structure of 4.
Scheme 3. Diels–Alder reaction of the transient species.
rearranged to 2a. From the literature it was known that the
,3-diene substructure of arene oxides can undergo Diels–
Alder reactions with dienophiles such as 7 to provide stable
1
[
6]
derivatives. By following this concept, 8 was isolated as a
single diastereomer and could even be characterized by X-ray
[
5]
crystal-structure analysis (Scheme 3, Figure 4).
1
Figure 2. H NMR spectrum of the transient species.
Under optimized conditions, 5a/6a can be enriched to
account for 80% of the material in the reaction mixture at
room temperature (Figure 3) and shows long-term stability
when cooled to À208C. At room temperature 5a/6a converts
directly into 2a (Scheme 3).
Two-dimensional NMR spectroscopy (H,H-COSY,
HMQC) of the reaction mixture at À208C strongly pointed
towards the presence of an arene oxide structure 5a. The
direct isolation of 5a failed, as the compound always
Figure 4. Crystal structure of 8.
[
7]
[8]
When other complexes, such as 9, 10 (X-ray crystal
[
5]
[8]
structure shown in Figure 5), 11, and even PtCl2 or
[
{(cod)IrCl} ] (cod = cyclooctadiene), were used, small tran-
2
1
sient peaks of 5a were also observed in the H NMR spectra
recorded during the reaction.
It is clear that 2a is much lower in energy than 1a. Density
functional calculations (B3LYP/6-31G** including zero-
point-energy (ZPE) correction) showed that even 5a and 6a
À1
are both 19 kcalmol lower in energy than 1a, and the
À1
reaction to give 2a then sets free another 42 kcalmol . On
the other hand, the experimentally observed relative energies
of 5a and 6a could not be reproduced at this level of theory:
In the NMR spectra, the signals for the diastereotopic
Figure 3. The amount y of the transient species as a percentage of the
material present in the reaction mixture is plotted against time. The
transient species can be accumulated to 80%. ^
2a,
&
1a,
~
5a/6a.
hydrogen atoms of both CH groups adjacent to the tosyl-
2
Angew. Chem. Int. Ed. 2005, 44, 2798 –2801
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2799

Communications
1
seems to be quite general; other substrates, such as 2b (R =
2
4
1
2
4
mesityl, R –R = H, X = NTs), 2c (R = Me, R –R = H, X =
1
2
4
1
NNs; Ns = nosyl), 2d (R = Me, R –R = H, X = O), 2e (R =
-Br-Ph, R –R = H, X = O), and 2 f (R = Me, R –R = H,
2
4
1
2
4
4
X =ÀCH(CH OÀallyl)OÀ) show the same chemical behav-
2
ior.
In conclusion, we have provided the first direct exper-
imental evidence for the formation of 2 via 5. The lack of a
primary kinetic isotope effect in the AuCl -catalyzed reac-
3
tions is an argument against pathway IV, by which, further-
more, the isomerization of the metalated arene oxide L to the
phenolate should be faster than a possible protodemetalation
of L. Thus, of the many conceivable pathways, only those that
proceed via 5, such as pathways I–III, can be responsible for
product formation. With a workup appropriate to the
sensitivity of 5, complex 4 might open a new entry to the
[
11]
entire chemistry of 5 from simple, readily available starting
materials. A further modification of the energy profile of the
reaction by ligand variation might reveal even more details of
the reaction mechanism.
Experimental Section
Synthesis of 8: 1a (45.0 mg, 150 mmol) was dissolved in CD CN, and 4
3
(2.70 mg, 7.50 mmol, 5 mol%) was added. The reaction was monitored
at room temperature by NMR spectroscopy. After five hours, the
solution was cooled to À408C, and 7 (26.4 mg, 150 mmol) was added.
The red solution was stored in the freezer at À258C for four days;
during this time it became colorless. After column chromatography of
the crude material (petroleum ether/ethyl acetate (PE/EA) 3:1), 8
(
31.3 mg, 44% from 1a) was obtained as a colorless solid. m.p. 1588C;
1
R (PE/EA 2:1): 0.23; H NMR (500 MHz, CD Cl ): d = 1.55 (s, 3H),
2
f
2
2
2
2
.38 (s, 3H), 3.32 (s, 1H), 3.55 (d, J = 11.6 Hz, 1H), 3.70 (dd, J =
4.6 Hz, 2.5 Hz, 1H), 4.00 (dd, J = 14.6 Hz, J = 2.0 Hz, 1H), 4.81 (d,
2
4
1
2
3
3
4
J = 11.6 Hz, 1H), 4.92 (d, J = 5.9 Hz, 1H), 5.96 (dt, J = 5.9 Hz, J =
[
a]
2
.25 Hz, 1H), 7.32–7.36 (m, 5H), 7.40–7.43 (m, 2H), 7.68 ppm (d,
Figure 5. Crystal strucuture of 10.
3
J = 8.2 Hz, 2H); [a] only a mean coupling constant could be
1
3
determined; C NMR (126 MHz, CD Cl ): d = 16.35 (q), 21.69 (q),
2
2
4
1
2
9.12 (t), 51.22 (d), 51.59 (t), 52.09 (q), 59.39 (d), 71.07 (s), 116.35 (d),
25.97 (d, 2C), 128.41 (d, 2C), 128.81 (d), 129.36 (d, 2C), 130.28 (d,
C), 131.57 (s), 132.37 (s), 138.96 (s), 144.96 (s), 154.64 (s), 155.36 ppm
amide group suggest that 5a is lower in energy than 6a and
that a possible equilibrium with low concentrations of 6a
must be slow on the NMR timescale; however, calculations
with different basis sets (B3LYP/6-31G**, BLYP/6-31G**, or
LMP2/6-31G**, each including ZPE correction) gave identi-
cal energies within the error of the method (a difference of
(s); IR (neat): n˜ = 1767, 1710, 1596, 1497, 1415, 1361, 1319, 1248, 1161,
À1
1095, 1076, 1063, 1025, 993, 811, 782, 760, 712, 693, 666 cm ; MS
+
(
7
FAB(+), 3-nitrobenzyl alcohol): m/z: 479 [M+H] ; MS (EI(À),
0 eV): m/z (%): 177 (42), 119 (64), 91 (100), 65 (22); C H N O S
24 22 4 5
(478.52).
À1 [9,10]
less than 1 kcalmol ).
Further support for an arene oxide structure was provided
Received: November 19, 2004
Revised: January 19, 2005
Published online: March 30, 2005
1
13
by the following characteristic H NMR and C NMR
spectroscopic data, which are in good agreement with
literature values for a substituted epoxide: The signal for
the hydrogen atom on the epoxide is a singlet at d =
Keywords: arene oxides · density functional calculations · gold ·
.
[
11]
1
[12]
homogeneous catalysis · N ligands
3
.87 ppm with J = 184 Hz; the tertiary and quaternary
CH
carbon atoms of the oxirane ring give rise to signals at d = 66.1
[
13]
and 69.5 ppm, respectively;
atoms on the methyl group is shifted to high field at d =
the signal for the hydrogen
[1] a) A. S. K. Hashmi, T. M. Frost, J. W. Bats, J. Am. Chem. Soc.
2000, 122, 11553 – 11554; b) A. S. K. Hashmi, T. M. Frost, J. W.
Bats, Org. Lett. 2001, 3, 3769 – 3771; c) A. S. K. Hashmi, T. M.
Frost, J. W. Bats, Catal. Today 2002, 72, 19 – 72; d) A. S. K.
Hashmi, L. Ding, J. W. Bats, P. Fischer, W. Frey, Chem. Eur. J.
3
1
.52 ppm, thus indicating an sp -hybridized neighboring
group.
Under the reaction conditions 5a does not undergo
interconversion into other constitutionally isomeric arene
2003, 9, 4339 – 4345; e) A. S. K. Hashmi, J. P. Weyrauch, M.
[
14]
oxides, as is known from the “oxygen walk” in the NIH shift
´
Rudolph, E. Kurpejovic, Angew. Chem. 2004, 116, 6707 – 6709;
Angew. Chem. Int. Ed. 2004, 43, 6545 – 6547.
[
15]
reaction.
The possibility of enriching the intermediate
2
800 ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2798 –2801

Angewandte
Chemie
[
2] a) B. Martꢀn-Matute, D. J. Cardenas, A. M. Echavarren, Angew.
Chem. 2001, 113, 4890 – 4893; Angew. Chem. Int. Ed. 2001, 40,
4754 – 4757; b) B. Martꢀn-Matute, C. Nevado, D. J. Cꢁrdenas,
A. M. Echavarren, J. Am. Chem. Soc. 2003, 125, 5757 – 5766.
3] a) E. Wenkert, E. L. Michelotti, C. S. Swindell, J. Am. Chem.
Soc. 1979, 101, 2246 – 2247; b) E. Wenkert, E. L. Michelotti, C. S.
Swindell, M. Tingoli, J. Org. Chem. 1984, 49, 4894 – 4899; c) E.
Wenkert, V. F. Ferreira, E. L. Michelotti, M. Tingoli, J. Org.
Chem. 1985, 50, 719 – 721; d) P. Kocienski, N. J. Dixon, S.
Wadman, Tetrahedron Lett. 1988, 29, 2357 – 2360.
[
[
4] L. Canovese, L. Cattalini, G. Marangoni, M. L. Tobe, J. Chem.
Soc. Dalton Trans. 1985, 731 – 735.
[
5] CCDC 256196 (4), 256266 (8), and 256267 (10) contain the
supplementary crystallographic data for this paper. These data
can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
6] a) M. P. Servꢂ, D. M. Jerina, J. Org. Chem. 1978, 43, 2711 – 2713;
b) J. R. Gillard, M. J. Newlands, J. N. Bridson, D. J. Burnell, Can.
J. Chem. 1991, 69, 1337 – 1343; c) K. Nojima, C. Isogami, M.
Hirobe, Chem. Pharm. Bull. 1993, 41, 2106 – 2108; d) A. P.
Henderson, E. Mutlu, A. Leclercq, C. Bleasdale, W. Clegg, R. A.
Henderson, B. T. Golding, Chem. Commun. 2002, 1956 – 1957.
7] L. S. Hollis, S. J. Lippard, J. Am. Chem. Soc. 1983, 105, 4293 –
[
[
[
4299.
8] A. Dar, K. Moss, S. M. Cottrill, R. V. Parisch, C. A. McAuliffe,
R. G. Pritchard, B. Beagley, J. Sandbank, J. Chem. Soc. Dalton
Trans. 1992, 1907 – 1913.
[
9] We thank Dr. G. Rauhut and Prof. Dr. H. Stoll, Institut fꢃr
Theoretische Chemie der Universitꢄt Stuttgart, for their con-
tinuous advice while we were carrying out the calculations.
[
10] In reference [2b], Echavarren and co-workers even suggest that
the oxepine is the ground state on the basis of PM3 calculations.
The difference in energy was estimated to be 10.3–12.7 kcal-
À1
mol
.
[
11] a) E. Vogel, H. Gꢃnther, Angew. Chem. 1967, 79, 429 – 446;
Angew. Chem. Int. Ed. Engl 1967, 6, 385 – 401; b) D. R. Boyd,
N. D. Sharma, Chem. Soc. Rev. 1996, 25, 289 – 296; c) L. I.
Belenꢅkii, Comp. Heterocycl. Chem. II 1996, 9, 45 – 66; d) S.
von Angerer, W. Tochtermann, Methoden Org. Chem. (Houben
Weyl) 4th ed. 1952–, Vol. E9d, 1998, pp. 1 – 64.
[
[
[
12] P. Laszlo, Bull. Soc. Chim. Fr. 1976, 1, 558 – 568.
13] H. Gꢃnther, G. Jikeli, Chem. Ber. 1973, 106, 1863 – 1880.
14] P. Y. Bruice, G. J. Kasperek, T. C. Bruice, H. Yagi, D. M. Jerina,
J. Am. Chem. Soc. 1973, 95, 1673 – 1674.
[
15] G. Guroff, J. W. Daly, D. M. Jerina, J. Renson, B. Witkop, S.
Udenfriend, Science 1967, 157, 1524 – 1530.
Angew. Chem. Int. Ed. 2005, 44, 2798 –2801
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2801