Solvent-controlled leaving-group selectivity in aromatic nucleophilic substitution

Article abstract of DOI:10.1021/ol801962a

(Figure Presented) A solvent-controlled inversion of leaving group ability allows selective access to either of two internal substitution products in S_NAr reactions of substrates with competing leaving groups. Application of this principle in a

Full text of DOI:10.1021/ol801962a

ORGANIC
LETTERS
2008
Vol. 10, No. 21
4859-4862
Solvent-Controlled Leaving-Group
Selectivity in Aromatic Nucleophilic
Substitution
Lukas Hintermann,† Ritsuki Masuo, and Keisuke Suzuki*
Department of Chemistry, Tokyo Institute of Technology, 2-12-1 O-okayama,
Meguro-ku, Tokyo 152-8551, Japan
ksuzuki@chem.titech.ac.jp
Received August 21, 2008
ABSTRACT
A solvent-controlled inversion of leaving group ability allows selective access to either of two internal substitution products in SNAr reactions
of substrates with competing leaving groups. Application of this principle in a selective synthesis of the highly functionalized xanthone core
of the antibiotic FD-594 is presented.
In nucleophilic aromatic substitution reactions (SNAr), a
range of kinetic studies, turned into textbook-knowledge, has
established the order of relative leaving-group ability: F, NO2
> Cl > Br > I, which is inverse to that in aliphatic
substitution.1,2 Alkoxides are good leaving groups only in
SNAr reactions,3 where OMe and Cl perform similarly,
though their relative ordering depends on solvent and
nucleophile.2a Competition between alkoxy and halogen
leaving groups in SNAr reactions is by no means esoteric;
we show here that the base-induced cyclization of 2-hy-
droxybenzophenones4 (1) with competing leaving groups X
and Y (X, Y ) halogen or alkoxide) in the 2′/6′-positions
gives xanthones 2 and 3 (Scheme 1). The ratio 2/3 reflects
Scheme 1. Intramolecular Competition of Leaving Groups X/Y
† Present address: Institut fu¨r Organische Chemie, RWTH Aachen
University, Germany.
(1) Smith, M. B.; March, J. March’s AdVanced Organic Chemistry, 5th
ed.; John Wiley: New York, 2001; p 860
.
(2) (a) Miller, J. Aromatic Nucleophilic Substitution; Elsevier: Amster-
dam, 1968. (b) Paradisi, C. ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon Press: Oxford, UK, 1991; Vol. 4, pp 423-
450.
the relative susceptibility for substitution under reaction
conditions.
(3) Use in synthesis: (a) Hattori, T.; Suzuki, T.; Hayashizaka, N.; Koike,
N.; Miyano, S. Bull. Chem. Soc. Jpn. 1993, 66, 3034–3040. (b) Reuman,
M.; Meyers, A. I. Tetrahedron 1985, 41, 837–860.
In planning synthetic strategies toward naturally occurring
oxygenated xanthones,5 competing modes of cyclization of
(4) (a) Hepworth, J. D. ComprehensiVe Heterocyclic Chemistry; Katritz-
ky, A. R., Ed.; Pergamon Press: Oxford, UK, 1984; Vol. 3, pp 835-840.
(b) Hepworth, J. D.; Gabbutt, C. D.; Heron, B. M. ComprehensiVe
Heterocyclic Chemistry II; Katritzky, A. R., Rees, C. W., Scriven, E. F. V.,
McKillop, A., Eds.; Elsevier: Oxford, UK, 1996; Vol. 5, p 351.
(5) Peres, V.; Nagem, T. J.; de Oliveira, F. F. Phytochemistry 2000, 55,
683–710.
10.1021/ol801962a CCC: $40.75
Published on Web 10/09/2008
2008 American Chemical Society

precursors 1 must be controlled. Here we present the concept
of solVent-dependent leaVing-group ability providing selec-
tively access to either xanthone 2 or 3 from a single precursor
1. Application of this principle to the synthesis of the highly
oxygenated xanthone core of the antibiotic FD-5946 show-
cases the synthetic utility of this approach.
Table 2. Competition of Chloride/Alkoxide Leaving Groupsa
A series of ortho-trisubstituted 2-hydroxybenzophenones
4a/b and 5a/b were obtained7 for investigating the relative
leaving-group ability of chloride or fluoride versus alkoxide
in several solvents8 (Tables 1 and 2). The fluorinated
entry
substrate solvent
T (°C)
80
time (h)
6/8
1
2
5a
5a
5a
5a
5a
5a
5b
5b
5b
5b
5b
MeOH
H2O
formamide80
THF
DMF
DMPU
MeOH
formamide80-90
THF
DMF
DMPU
75
27
30
30
5
2.3
20
70
30
3
1:11b
1:8.6
1:7.7
1:8.5
100
3c
4d
5
Table 1. Competition of Fluoride/Alkoxide Leaving Groupsa
80
45
1.7:1e
6
7
50
2.8:1
100f
1:8g
8h
9d
10
11
1:3.4
1:5.5
10:1
20:1
80
50
50
5
a General conditions: see Table 1. b 98% yield (90% 8 and 8% 6a).
c 97% conversion. d Heterogeneous reaction mixture in THF. e 99% yield
(62% 6a and 37% 8). f Reaction in a sealed tube. g In addition, 6a (3%)
was also formed. h Conversion ) 53%.
entry substrate solvent T (°C) time (h)
6/7
yield (%)
1
2
3
4
4a
4a
4b
4b
MeOH
DMF
MeOH
DMF
85
50
85
rt
1.4
0.7
1.3
1
100:0
100:0
100:0
100:0
99
99
99
nd
a General conditions for all tables: 10 mg of substrate/mL solvent; 1.5
equiv of Cs2CO3. T ) reaction temperature; time ) reaction time to
complete conversion (TLC), unless otherwise indicated. Product ratios are
Scheme 2 shows a synthetic approach to xanthone 9: ether
protection and desymmetrization of terephthalate 1011 by
saponification was followed by transformation to aldehyde
1
derived from H NMR of the crude mixture.
benzophenones 4a/b gave the corresponding 1-alkoxyxan-
thones 6a/b in almost quantitative yield with no trace of
fluoroxanthone 7, regardless of the alkoxy group or the
solvent (Table 1).9
On the other hand, the chlorinated substrates 5 cyclized
to give mixtures of alkoxyxanthones 6 and chloroxanthone
8 (Table 2). The product ratios are subject to a substantial
solVent effect, which causes a remarkable reversal of the
product selectivity. The magnitude of selectivity reversal is
more pronounced for isopropyl ether 5b (entry 7 vs 11) than
for methyl ether 5a (entry 1 vs 6).
Figure 1. Structure of FD-594 and a derived xanthone core 9.
The antibiotic FD-5946 contains a highly substituted
oxygenated xanthone core (the DEF rings), which we chose
to disconnect to a key intermediate 9 for a projected total
synthesis (Figure 1).10
11. Addition of lithioarene 12 to aldehyde 11 followed by
oxidation/deprotection gave the cyclization precursor 14.12
The cyclization of 14 with cesium carbonate as a base in
different solvents gave a variable mixture of 1,4-dialkoxyx-
anthone 15 and 1,4-dichloroxanthone 16 (Table 3). The
solvent effect on the leaving-group selectivity was even more
pronounced than for the model compounds 5. Either es-
sentially pure 16 (entries 1-4) or 15 (entries 9 and 12) was
obtained in suitable solvents. The reactions that are selective
(6) (a) Qiao, Y. F.; Okazaki, T.; Ando, T.; Mizoue, K. J. Antibiot. 1998,
51, 282–287. (b) Kondo, K.; Eguchi, T.; Kakinuma, K.; Mizoue, K.; Qiao,
Y. F. J. Antibiot. 1998, 51, 288–295. (c) Eguchi, T.; Kondo, K.; Kakinuma,
K.; Uekusa, H.; Ohashi, Y.; Mizoue, K.; Qiao, Y.-F. J. Org. Chem. 1999,
64, 5371–5376.
(7) These model compounds were prepared by (a) addition of metalated
3-alkoxyhalobenzenes to O-MOM-salicylaldehyde, (b) oxidation of the
intermediary diarylmethanols, and (c) MOM-deprotection; see the Support-
ing Information.
(8) Solvent effects on leaving-group order in SNAr reactions have been
documented through kinetic measurements; see ref 2a.
M.; Kitamura, M.; Kato, H.; Oorui, M.; Suzuki, K. Angew. Chem., Int. Ed.
2005, 44, 3871–3874. (c) Kitamura, M.; Ohmori, K.; Kawase, T.; Suzuki,
K. Angew. Chem., Int. Ed. 1999, 38, 1229–1232.
(9) In case of reaction in methanol, this result is in accordance to an
earlier report: Horne, S.; Rodrigo, R. J. Org. Chem. 1990, 55, 4520–4522.
(10) This disconnection is based on the synthetic strategy already used
for the total syntheses of the benanomicin-pradimicin antibiotics: (a)
Tamiya, M.; Ohmori, K.; Kitamura, M.; Kato, H.; Arai, T.; Oorui, M.;
Suzuki, K. Chem.sEur. J. 2007, 13, 9791–9823. (b) Ohmori, K.; Tamiya,
(11) Hintermann, L.; Suzuki, K. Synthesis 2008, 2303–2306.
(12) See the Supporting Information for details of this synthesis.
(13) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry,
2nd ed.; VCH: Weinheim, Germany, 1988.
4860
Org. Lett., Vol. 10, No. 21, 2008

dipolar aprotic solvents with strong donor capabilities cluster
together in a separate region (Figure 2).
Scheme 2.
Synthesis of the FD-594 Xanthone Core12
Figure 2. Product selectivity in the reaction 14 f15 + 16 as a
function of solvent polartiy. Data points are the entries of Table 3.
for 16 are slower than those leading primarily to 15 (compare
entry 3 vs 9). The presence of the radical scavengers 2,6-
di-tert-butylcresole or sodium meta-nitrobenzenesulfonate did
not affect the outcome in DMF. The reaction in toluene was
unaffected by the presence of either [Cu(MeCN)4]PF6 or
Pd(OAc)2/dppf.
The mechanism of nucleophilic aromatic substitution (SNAr)
includes two steps, namely, the nucleophilic attack on the
aromatic system (transition state T1) and extrusion of the leaving
group (T2).2a For moderate nucleophiles such as those centered
on oxygen or nitrogen, attack to the aromatic ring (T1) is usually
rate-limiting, which explains the curious leaving-group order:
F, OR > Cl > Br, paralleling the electrophilicity of the ipso-
carbon to welcome the incoming nucleophile.1,2
In reactions with polarizable, kinetically fast nucleophiles,
T2 is rate-determining,2a and the classical leaving-group order
is observed (Cl > F > OR). In cyclizations to xanthones, the
entropically favored nucleophilic attack may outweigh the low
nucleophilicity of phenolate, such that T1 and T2 become similar
in energy. Dipolar aprotic solvents increase the nucleophilicity
of phenoxide; T1 is further lowered in energy, T2 becomes rate-
limiting, and we observe fast reactions with high selectivity for
the halogen-leaving groups (Cl, F > OR). In other solvents,
increasing solvent polarity will decrease the phenoxide nucleo-
philicity by solvation (increasing T1) and better solvate the
leaving group (decreasing T2). This simple picture explains the
observed reversal of the leaVing group order in the cyclization
of 2-chloro-6-alkoxy-2′-hydroxybenzophenones as a change of
the rate-determining step from T1 (polar solvents) to T2
(nonpolar solvents).
When the chemoselectivity of cyclization (measured as
logarithm of the product ratio 15/16) is plotted as a function
13
N
of solvent polarity ET (30), the data points for a range of
solvents lie approximately on a straight line, but several
Table 3. Cyclization of 14 to 15/16 in a Range of Solventsa
entry
solvent
ETN(30)b T (°C) time (h)
15/16
1
2
3
4
5
6
7
8
CF3CH2OHc
EtOH (aq)d
formamide
MeOH
0.898
0.885e
0.799
0.762
0.762
0.60 g
0.460
0.444
0.404
0.404
0.355
0.352
0.321
0.302
0.244
0.207
0.099
0.006
80
80
50
80
80
50
50
rt
70
7.5
24
14
13
2.3
2
8
0.5
1:50
1:27.5
1:25
1:22
1:23
1:2.9
3.2:1
12:1
19:1
16:1
5:1
25:1
1:1.95
2:1
3.2:1
1.65:1
2.5:1
8:1
MeOD
[bmim]PF6
f
MeCN
DMSO
DMF
9
45
rt
10
11
12
13
14
15
16
17
18
DMF
<20
acetone
DMPU
tAmOH
pyridine
diglyme
THF
50
50
90
50
50
80
130
90
2
0.7
18
2
3
1.7
2.3
Fluorinated substrates do not show this ambiguity because
fluorine is a superior leaving group over alkoxide in both
transition states. Choice of fluorine as the leaving group
ensures selective ring closure, but this option will often be
out of reach due to the difficulty of introducing aromatic
fluorine into advanced synthetic intermediates.15
toluene
c-C6H12
We conclude that, in synthetic routes where alkoxy and
chlorine leaving groups compete for an internal nucleophile
18
a General conditions: see Table 1. b Solvent polarity parameter.13
c 6 equiv of Cs2CO3, conversion ) 40%. d H2O/EtOH ) 2:1. e Interpolated
value. f N-Butyl-N′-methylimidazolium hexafluorophosphate. g Value (ET(30)
≈ 50 kcal/mol) calculated from lit.14 diglyme ) O(C2H4OMe)2; DMPU )
(14) Aki, S. N. V. K.; Brennecke, J. F.; Samanta, A. Chem. Commun.
2001, 413–414.
t
N,N′-dimethylpropylidene urea; AmOH ) 2-methyl-2-butanol.
(15) Hudlicky, M.; Pavlath, A. E. Chemistry of Organic Fluorine
Compounds II; American Chemical Society: Washington, DC, 1995.
Org. Lett., Vol. 10, No. 21, 2008
4861

in an SNAr reaction, the relative leaving-group order can be
controlled by a solvent effect. The finding is relevant to the
synthesis of highly functionalized and nonsymmetrically
substituted xanthones and may be of more general importance
in the case of SNAr reactions with competing leaving groups.
Society for the Promotion of Science (JSPS) for a postdoc-
toral fellowship.
Supporting Information Available: Experimental pro-
cedures and substance characterization data. This material
is available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. This work is dedicated to the memory
OL801962A
of the late Charles Mioskowski. L.H. thanks the Japan
4862
Org. Lett., Vol. 10, No. 21, 2008