A highly active ytterbium(III) methide complex for truly catalytic Friedel-Crafts acylation reactions

Article abstract of DOI:10.1055/s-2002-34231

The Friedel-Crafts acylation of anisole with acetic anhydride using ytterbium(III) tri[tris(nonafluorobutanesulfonyl)methide] was studied with respect to catalyst loading. A strong inhibitory effect due to the product became apparent from doping experiments and from examination of the kinetic data. This understanding allowed catalyst loadings to be reduced to as little as 0.1 mol% for effective acylation under a suitable temperature and pressure regime.

Full text of DOI:10.1055/s-2002-34231

LETTER
1653
A Highly Active Ytterbium(III) Methide Complex for Truly Catalytic
Friedel–Crafts Acylation Reactions
a
a
a
b
a
A
Ytterbium(III)
M
n
ethide
C
om
t
plex
f
o
h
r
TrulyCata
o
F
riedel–
n
Crafts
A
cyla
y
tion Reactions G. M. Barrett,* Nathalie Bouloc, D. Christopher Braddock, David Chadwick, David A. Henderson
a
Department of Chemistry, Imperial College of Science, Technology and Medicine, South Kensington, London SW7 2AY, UK
E-mail: agmb@ic.ac.uk
b
Department of Chemical Engineering, Imperial College of Science, Technology and Medicine, South Kensington, London SW7 2AY,
UK
Received 15 August 2002
um complexes, which operate at low catalyst loadings (2.5
mol%) for Friedel–Crafts acylation reactions.5
Abstract: The Friedel–Crafts acylation of anisole with acetic anhy-
dride using ytterbium(III) tri[tris(nonafluorobutanesulfonyl)me-
thide] was studied with respect to catalyst loading. A strong We have recently introduced ytterbium(III) and scandi-
inhibitory effect due to the product became apparent from doping
um(III) tris(trifluoromethanesulfonyl)methides (‘tri-
experiments and from examination of the kinetic data. This under-
flides’) Yb[C(SO CF ) ] and6 Sc[C(SO CF ) ] as
2
3 3 3
2
3 3 3
standing allowed catalyst loadings to be reduced to as little as 0.1
mol% for effective acylation under a suitable temperature and pres-
sure regime.
catalysts for aromatic nitration. The triflide catalysts
were found to be far superior than the corresponding tri-
flate or triflamide salts. In that work we reported a novel
route to the triflide ligand, the preparation of the ytterbium
and scandium catalysts, X-ray crystal structures of both
and their repeated recycling and re-use for nitration.
These catalysts are also excellent for the nitration of per-
fluoroaromatics where complications with HF liberation
Key words: Friedel–Crafts, acylation, catalytic, ytterbium(III)
methide, inhibition
Friedel–Crafts acylation reactions of arenes are of funda-
mental importance for the synthesis and manufacture of
7
do not arise. We have also disclosed the preparation of a
series of ytterbium methide salts Yb[C(SO R ) ] with
1
aromatic ketones. These reactions are classically per-
2
f3 3 3
formed by making use of strong, oxophilic Lewis acids
such as aluminum trichloride, titanium tetrachloride or
tin(IV) chloride. However, at least stoichiometric quanti-
ties of the Lewis acid are required since the product ke-
tone sequesters the Lewis acid ‘catalyst’ rendering it
inactive. Moreover, this Lewis acid–ketone complex must
be destroyed by an aqueous quench after completion of
the Friedel–Crafts reaction leading to serious problems
with strongly acidic waste streams. The need for truly cat-
alytic acylation with recoverable and recyclable catalysts
remains strong. We now report on the use of a lan-
thanide(III) methide complex that functions efficiently at
low catalytic loadings (0.1–1 mol%) under appropriate
pressure and temperature regimes.
long chain-perfluorinated ponytails (R = C F , 1; C F ;
f
4
9
6 13
C F ) leading ultimately to a fluorous biphase acylation
8
17
system where the catalyst can be recycled and re-used
without loss in activity or yield.8 Mikami has also ap-
plied these catalyst types for Friedel–Crafts acylation re-
,9
10
actions. However, both in our work and in Mikami’s, a
0 mol% catalyst loading was employed. Due to the rela-
1
tively high molecular weights of these complexes, com-
pared to the reactants, it would be desirable to use much
reduced levels of catalyst.11
Much research has been performed in an effort to procure
an effective Friedel–Crafts acylation catalyst. Kobayashi
Scheme 1
has demonstrated the catalytic use of lanthanide and
2
hafnium triflate salts [M(OTf) ; n = 3, 4] for Friedel–
n
Under our previously described standard conditions (10
Crafts acylation reactions and has shown the catalyst may
mol% 1, 2 equiv Ac O, trifluoromethylbenzene solvent,
2
be recycled in selected instances. Catalytic acylation with
8
2
5 °C) anisole was acylated to 69% conversion in 3 hours
3
bismuth(III) triflate
and ytterbium triflamide
(
Scheme 1). We were encouraged to find that decreasing
4
(
Yb(NTf ) ) have also been reported. Typically in these
2 3
the catalyst loading to 1 mol% 1 gave a 41% conversion
under identical conditions. Moreover, at just 0.1 mol% 1
catalyst loading a 17% conversion to the aromatic ketone
products was observed. The non-linear correlation be-
tween catalyst loading and conversion suggested complex
kinetic behaviour and accordingly we undertook a more
detailed study. When these acylation reactions were mon-
itored from t = 0 it was found that the initial rates were
very fast but then slowed rapidly with increasing conver-
sion. This behaviour is typical of product inhibition. In
systems a 10 mol% catalyst loading is required for satis-
factory yields. More recently, in work that takes advan-
tage of a product-catalyst ‘hard-soft’ mismatch to
facilitate turnover Fürstner has reported cationic palladi-
Synlett 2002, No. 10, Print: 01 10 2002.
Art Id.1437-2096,E;2002,0,10,1653,1656,ftx,en;D30302ST.pdf.
©
Georg Thieme Verlag Stuttgart · New York
ISSN 0936-5214

1
654
A. G. M. Barrett et al.
LETTER
Table 2 Influence of Temperature and Catalyst Loadinga
principle, this inhibitory effect could arise from two
sources: complexation of the de facto catalytic species
with the aromatic ketone product as it is generated, or
from the acetic acid by-product as the anhydride is con-
sumed. To test this notion, acylation runs were performed
on anisole, doping the mixtures with either 100 mol% ace-
tic acid or 100 mol% of representative ketones (Table 1).
While the inhibitory effect of acetic acid (entry 2) was
found to be minimal, the effect of added ketones was dra-
matic. The strongly electron deficient 4-nitroacetophe-
none (entry 5) had a negligible effect on the conversion
rate, but doping with the electron-rich 2,4-dimethoxyace-
tophenone (entry 3) completely inhibited the reaction. Ac-
etophenone itself (entry 4) showed an intermediate effect
as might be expected on electronic grounds. It is therefore
apparent that the inhibition effect is due to the aromatic
ketone product.
% Conversionb
mol% 1
Temp (°C)
1
0
1
1
1
1
0
25
25
50
70
90
25
50
70
90
25
50
100
64 (81)
34 (54)
61 (72)
65 (71)
73 (79)
16 (19)
32 (40)
34 (32)
39 (37)
21 (26)
37 (49)
59 (64)
.1
0.1
0
0
0
0
0
.1
Table 1 Inhibition Studies
.1
.1c
.1c
.1c
Entry
Dopanta
% Conversionb
69 (74)
1
2
3
4
–
a
All reactions were perfomed in trifluoromethylbenzene solvent
using 2 equiv of anhydride for 1.5 h with catalyst 1.
acetic acid
64 (78)
b
As determined by GC/MS analysis. Figures in parentheses refer to
2,4-dimethoxyacetophenone
acetophenone
4-nitroacetophenone
0 (0)
conversion after 20 h.
c
No solvent (reaction performed neat).
36 (60)
5
57 (75)
a
1
00 mol%.
b
Determined by GC/MS analysis after 3 h. Figures in parentheses
refer to conversion after 20 h.
From the above, it became apparent that to increase cata-
lytic competence it is imperative to discourage strong cat-
alyst–product interaction (i.e. encourage catalyst–product
dissociation). Based on simple entropic considerations, an
increase in temperature might be expected to suppress
such complex formation. Accordingly, we examined the
effect of temperature on the acylation of anisole with ace-
tic anhydride using 1 at 10, 1 and 0.1 mol% loading up to
the boiling point of 102 °C of trifluorobenzene (Table 2).
The runs performed with 1 mol% 1 in BTF (25 °C, 50 °C,
Figure 1
0 °C) are also displayed graphically (Figure 1).12
after 1.5 hours. This result clearly demonstrates that low
catalyst loadings can be employed in Friedel–Crafts acy-
lation reactions by proper consideration of the nature and
thermal stability of the inhibitory ketone–catalyst com-
plex.
7
It is clear from the results presented in Table 2 and
Figure 1 that increasing the temperature increases the
yield dramatically for a given catalyst loading, supporting
our postulate of product inhibition. For example, the use
of 1 mol% 1 at 90 °C (73%, 1.5 h) is more effective than
the use of 10 mol% 1 at 25 °C (64%, 1.5 h), and when run
neat at 100 °C the use of 0.1 mol% 1 is competitive (59%,
The efficacy of Friedel–Crafts acylation reactions of var-
ious arenes with acetic, isobutyric and benzoic anhydrides
catalyzed by 1 mol% 1 under these conditions were exam-
ined (Table 3).1 Inspection of the results shows that the
reaction is general, although the benzoic anhydride is a su-
perior acylating agent compared to the aliphatic anhy-
drides. As expected, less electron rich arenes undergo
3,14
1
.5 h). An extrapolation to higher temperatures would
suggest further improvement. When heated to 150 °C in a
sealed tube we were pleased to find that at 1 mol% loading
of 1 a 90% conversion to ketone products was observed
Synlett 2002, No. 10, 1653–1656 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
An Ytterbium(III) Methide Complex for Truly Catalytic Friedel–Crafts Acylation Reactions
1655
Table 3 Catalytic Friedel–Crafts Acylation Reactionsa
Acknowledgement
We thank the EPSRC and the Institute of Applied Catalysis (iAc)
for support of our research, Glaxo Group Research Ltd. for the most
generous endowment (to A.G.M.B.) and the Wolfson Foundation
for establishing the Wolfson Centre for Organic Chemistry in Me-
dical Science at Imperial College. We thank Prof. Chris Adams
ArH
R
% Yieldb
80 (90, 92)
90 (97, 98)
91 (99, 99)
33 (26, 37)
43 (43, 46)
89 (95, 99)
79c
Isomer ratiod
97% para only
para only
para only
para only
para only
para only
99% 2-
(
iAc) and Dr Chris Newmann (Quest International) for useful dis-
Anisole
Me
cussions.
Anisole
(CH ) CH-
3
2
References
Anisole
Ph
(1) (a) Olah, G. A. Friedel–Crafts Chemistry; Wiley-
Interscience: London, 1973. (b) Heaney, H. In
Thioanisole
Thioanisole
Thioanisole
Thiophene
Thiophene
Thiophene
Me
Comprehensive Organic Synthesis, Vol. 2; Trost, B. M.;
Fleming, I., Eds.; Pergamon Press: Oxford, 1991, 733.
2) (a) Kawada, A.; Mitamura, S.; Kobayashi, S. J. Chem. Soc.,
Chem. Commun. 1993, 1157. (b) Kawada, A.; Mitamura, S.;
Kobayashi, S. Synlett 1994, 545. (c) Hachiya, I.; Moriwaki,
M.; Kobayshi, S. Bull. Chem. Soc. Jpn. 1995, 68, 2053.
(d) Kobayashi, S.; Iwamoto, S. Tetrahedron Lett. 1998, 39,
(CH ) CH-
3
2
(
Ph
Me
(CH ) CH-
45c
2- only
3
2
4697.
Ph
49c
2- only
(
3) (a) Desmur, J. R.; Labrouillere, M.; Le Roux, C.; Gaspard,
66c
H.; Laporterie, A.; Dubac, J. Tetrahedron Lett. 1997, 38,
3
3
3
-Methylthiophene Me
-Methylthiophene (CH ) CH-
2-:5- 73:27
2-:5- 60:40
2-:5- 65:35
9- only
8871. (b) Repichet, S.; LeRoux, C.; Dubac, J.; Desmurs, J.
88c
R. Eur. J. Org. Chem. 1998, 2743.
3
2
(
(
(
4) Mikami, K.; Kotera, O.; Motoyama, Y.; Sakaguchi, H.;
Maruta, M. Synlett 1996, 171.
5) Fürstner, A.; Voightlander, D.; Schrader, W.; Giebel, D.;
Reetz, M. T. Org. Lett. 2001, 3, 417.
6) Waller, F. J.; Barrett, A. G. M.; Braddock, D. C.;
Ramprasad, D.; McKinnell, R. M.; White, J. P.; Williams, D.
J.; Ducray, R. J. Org. Chem. 1999, 64, 2910.
7) Barrett, A. G. M.; Braddock, D. C.; Ducray, R.; McKinnell,
R. M.; Waller, F. J. Synlett 2000, 57.
8) (a) Barrett, A. G. M.; Braddock, D. C.; Catterick, D.;
Chadwick, D.; Henschke, J. P.; McKinnell, R. M. Synlett
2000, 847. (b) Barrett, A. G. M.; Bouloc, N.; Braddock, D.
C.; Catterick, D.; Chadwick, D.; White, A. J. P.; Williams,
D. J. Tetrahedron 2002, 58, 3835.
-Methylthiophene Ph
56c
Anthracene
Naphthalene
m-Xylene
Toluene
Ph
Ph
Ph
Ph
Ph
65 (64, 80)
93 (90, 99)
81 (69, 87)
66 (33, 73)
0 (0, 0)
: 37:63
ortho only
15:<1: 84e
–
(
(
Chlorobenzene
a
All reactions were performed in trifluoromethylbenzene solvent us-
ing 2 equiv of anhydride at 150 °C for 20 h with 1 mol% 1 in a sealed
tube.
Isolated yield after chromatography. Figures in parentheses refer to
conversion after 1.5 h and then 20 h as determined by GC/MS.
b
(9) Catalytic Compounds and Processes. UK Patent Application
No 9919583.6.
c
Experiments conducted at 80 °C.
Isomer ratio after chromatography.
o:m:p.
(10) (a) Mikami, K.; Mikami, Y.; Matsumoto, Y.; Nishikido, J.;
Yamamoto, F.; Nakajimi, H. Tetrahedron Lett. 2001, 42,
289. (b) Mikami, K.; Mikami, Y.; Matsuzawa, H.;
Matsumoto, Y.; Nishikido, J.; Yamamoto, F.; Nakajima, H.
Tetrahedron 2002, 58, 4015.
d
e
(
11) For experimental details for the preparation of these
complexes see: Barrett, A. G. M.; Bouloc, N.; Braddock, D.
C.; Catterick, D.; Chadwick, D.; White, A. J. P.; Williams,
D. J. Tetrahedron 2002, 58, 3835.
acylation more slowly than anisole. However, we note
that these conditions allow the acylation of toluene with
anhydrides, rather than acid chlorides, under lanthanide
catalysis for the first time.
(12) These plots were obtained using a Mettler-Toledo REACT-
IR 4000 system with a dicomp probe scanning every 60 s
In conclusion, we have demonstrated that ytterbium(III)
methide complex 1 is an effective catalyst for Friedel–
Crafts acylation at significantly reduced catalyst loadings.
The catalyst is, like traditional Lewis acid catalyst, subject
to product inhibition. However, moderate heating appar-
ently leads to significant product–catalyst dissociation
and thereby facilitates catalytic turnover. To the best of
our knowledge these low catalyst loadings are the best re-
ported to date for Friedel–Crafts acylation reactions.
(
25 °C, 50 °C) or 45 s (70 °C) for the first two hours followed
by one scan every five min for the remaining time. The
profiles are generated by following the peak area between
1
672 and 1685 wave numbers to a constant baseline point.
(
13) General Experimental Procedure: A stirred solution of
arene (0.6 mmol), anhydride (1.2 mL) and catalyst 1 (16 mg,
0
.006 mmol) in PhCF (1.2 mL) was heated to 150 °C in a
3
sealed tube for 20 h. The resulting mixture was allowed to
cool, diluted with Et O (10 mL), washed with water (2 10
mL), sat. aq sodium hydrogen carbonate solution (2 10
mL), sat. aq sodium chloride solution (2 10 mL), dried
2
(
MgSO ), concentrated in vacuo and chromatographed.
4
Synlett 2002, No. 10, 1653–1656 ISSN 0936-5214 © Thieme Stuttgart · New York

1
656
A. G. M. Barrett et al.
LETTER
1
(
14) IR and H NMR for Friedel–Crafts Acylation Products:
J = 3.5 Hz), 7.52–7.68 (m, 6 H), 7.87 (d, 2 H, J = 3.0 Hz). 5-
–
1
1
–1 1
4
(
-Methoxyacetophenone: IR (KBr): 1676 cm . H NMR
300 MHz, CDCl3): = 2.58 (s, 3 H), 3.84 (s, 3 H), 6.96 (d,
H, J = 9.0 Hz), 7.95 (d, 2 H, J = 9.0 Hz). 4-Methoxy-
isobutyrophenone: IR (KBr): 1675 cm . H NMR (300
MHz, CDCl3): = 1.20 (d, 6 H, J = 7.0 Hz), 3.53 (septet, 1
H, J = 7.0 Hz), 3.87 (s, 3 H), 6.94 (d, 2 H, J = 8.5 Hz), 7.95
Acetyl-3-methylthiophene: IR (KBr): 1662 cm . H NMR
(270 MHz, CDCl3): = 2.29 (s, 3 H), 2.53 (s, 3 H), 7.22 (d,
1 H, J = 1.0 Hz), 7.49 (d, 1 H, J = 1.0 Hz). 2-Acetyl-3-
2
–
1
1
–1 1
methylthiophene: IR (KBr): 1662 cm . H NMR (270
MHz, CDCl3): = 2.29 (s, 3 H), 2.55 (s, 3 H), 6.93 (d, 1 H,
J = 5.0 Hz), 7.39 (d, 1 H, J = 5.0 Hz). 2- and 5-Isobutyryl-
–
1 1
(
d, 2 H, J = 8.5 Hz). 4-Methoxybenzophenone: IR (KBr):
3-methylthiophene: IR (KBr): 1660 cm . H NMR (300
MHz, CDCl3): = 1.23 (d, 6 H, J = 7.0 Hz), 2.13 (s, 1.05 H),
2.59 (s, 1.95 H), 3.32 (m, 1 H), 6.97 (d, 0.65 H, J = 5.0 Hz),
7.38 (s, 0.35 H), 7.40 (d, 0.65 H, J = 5.0 Hz), 7.55 (s, 0.35
H). 2- and 5-Benzoyl-3-methylthiophene: IR (KBr): 1633
–
1 1
1
6
7
650 cm . H NMR (300 MHz, CDCl ): = 3.91 (s, 3 H),
3
.98 (d, 2 H, J = 8.0 Hz), 7.47–7.59 (m, 3 H), 7.78 (d, 2 H,
.0 Hz), 7.85 (d, 2 H, J = 8.0 Hz). 4-Thiomethylace-
tophenone: IR (KBr): 1673 cm . H NMR (300 MHz,
–
1 1
–
1 1
CDCl ): = 2.55 (s, 3 H), 2.59 (s, 3 H), 7.28 (d, 2 H, J = 8.0
cm . H NMR (300 MHz, CDCl ): = 2.34 (s, 1.2 H), 2.51
3
3
Hz), 7.89 (d, 2 H, J = 8.0 Hz). 4-Thiomethyl-isobutyro-
phenone: IR (KBr): 1670 cm . H NMR (300 MHz,
(s, 1.8 H), 7.03 (d, 0.6 H, J = 5.0 Hz), 7.34 (s, 0.4 H), 7.46–
7.61 (m, 4 H), 7.86 (m, 2 H). 9-Benzoylanthracene: IR
–
1 1
–
1 1
CDCl ): = 1.21 (d, 6 H, J = 7.0 Hz), 2.53 (s, 3 H), 3.51
(KBr): 1656 cm . H NMR (300 MHz, CDCl ): = 7.39–
3
3
(
septet, 1 H, J = 7.0 Hz), 7.27 (d, 2 H, J = 8.5 Hz), 7.87 (d, 2
7.61 (m, 7 H), 7.74 (d, 2 H, J = 8.5 Hz), 7.84 (d, 2 H, J = 7.0
Hz), 8.09 (d, 2 H, J = 8.5 Hz), 8.56 (s, 1 H). - and - Benz-
oylnaphthalene: IR (KBr): 1659 cm . H NMR (300 MHz,
H, J = 8.5 Hz). 4-Thiomethylbenzophenone: IR (KBr):
–1
1
–1 1
1
7
2
649 cm . H NMR (300 MHz, CDCl ): = 2.56 (s, 3 H),
3
.28–7.32 (2 H, m), 7.48–7.65 (3 H, m), 7.72-7.80 (m, 4 H).
CDCl ): = 7.46–7.65 (m, 7 H), 7.88–8.05 (m, 4 H), 8.11 (d,
3
–
1 1
-Acetylthiophene: IR (KBr): 1662 cm . H NMR (270
0.65 H, J = 7.5 Hz), 8.29 (s, 0.35 H). 2,4-Dimethyl-
–
1 1
MHz, CDCl3): = 2.59 (s, 3 H), 7.15 (1 H, t, J = 4.0 Hz) 7.65
d, 1 H, J = 4.0 Hz), 7.71 (d, 1 H, J = 4.0 Hz). 2-Isobutyryl-
benzophenone: IR (KBr): 1662 cm . H NMR (300 MHz,
(
CDCl ): = 2.35 (s, 3 H), 2.41 (s, 3 H), 7.07 (d, 1 H, J = 7.0
3
–
1 1
thiophene: IR (KBr): 1662 cm . H NMR (300 MHz,
CDCl ): = 1.27 (d, 6 H, J = 7.0 Hz), 3.41 (septet, 1 H,
J = 7.0 Hz) 7.15 (t, 1 H, J = 4.5 Hz) 7.65 (d, 1 H, J = 5.0 Hz),
Hz), 7.13 (s, 1 H), 7.25 (d, 1 H, J = 7.5 Hz), 7.47 (m, 2 H),
7.58 (m, 1 H), 7.81 (d, 2 H, J = 7.0 Hz). 2- and 4-Methyl-
3
–
1 1
benzophenone: IR (KBr): 1658 cm . H NMR (300 MHz,
7
.75 (d, 1 H, J = 4.5 Hz). 2-Benzoylthiophene: IR (KBr):
CDCl ): = 2.36 (s, 0.4 H, 2-CH ), 2.45 (s, 2.6 H, 4-CH ),
7.29–7.82 (m, 9 H).
3
3
3
–1 1
1632 cm . H NMR (300 MHz, CDCl ): = 7.19 (t, 1 H,
3
Synlett 2002, No. 10, 1653–1656 ISSN 0936-5214 © Thieme Stuttgart · New York