Angewandte
Communications
Chemie
ion, a catalytic enantioselective route to spiro b-dicarbonyl
derivatives could be envisaged. Spiro compounds are valuable
intermediates in synthesis, materials and catalysis,[14] but
catalytic enantioselective routes for their synthesis are
relatively underdeveloped.[15] Catalytic enantioselective
methods for the generation of axially chiral spiro compounds
are particularly rare.[16]
We prepared a suitable substrate 7 in which the acyl group
was activated as a phenyl ester in three steps from indan-1-
one.[17] Upon exposure of 7 to tetrabutylammonium bromide
(TBAB) and 50% aqueous potassium carbonate in toluene at
room temperature, we observed smooth cyclization to the
racemic C2-symmetric spirobiindanone 9, with no trace of O-
acylation observed (Table 1, entry 1).
Use of N-benzyl cinchonidinium chloride 10 in place of
TBAB gave complete conversion to the desired product and
a modest e.r. of 56:44. We observed that the use of aqueous
hydroxide bases led to hydrolysis of the phenyl ester 7
(entry 3) and hence we switched to solid bases in order to
screen catalysts for higher enantioselectivity. This led to
a small increase in selectivity (to 59:41 e.r., entry 4) and under
these conditions no hydrolysis was observed. We subse-
quently screened a range of ammonium salts and under these
conditions, 11, which bears a urea as a Brønsted acidic group,
was found to give improved selectivity of 69:31 e.r. (entry 5).
We were concerned that the relatively low selectivities
observed were a consequence of a background reaction
mediated by the basic phenolate leaving group, and hence we
examined a reaction with a stoichiometric amount of sodium
phenolate in the presence of catalyst 11 (entry 6). This led to
very low enantioselectivity (51: 49 e.r.), which was consistent
with our hypothesis. Consequently, we changed our substrate
to pentafluorophenyl ester 8, rationalizing that the lower pKa
of pentafluorophenol to phenol (19.5 vs. 27.4 in MeCN)[18]
would limit this background reaction.[19] Subjecting this
substrate 8 to the same conditions (entry 7) led to a relatively
poor 42:58 e.r. but a significantly slower reaction time, and
hence we decided to reoptimize the catalyst for this sub-
strate.[17] Changing to quininium catalyst 12 led to a signifi-
cantly higher selectivity (79:21 e.r., entry 8), and pseudo-
enantiomer 13 led to a very similar 25:75 e.r. and the expected
reversal of absolute configuration (entry 9). We subsequently
examined a range of different N-substituents on the quinidi-
nium nucleus; N-pentafluorophenylbenzyl catalyst 14 gave
poorer selectivity (36:64 e.r., entry 10) but we observed that
selectivity improved with larger N-substituents (29:71 e.r.
with 3,5-di-tert-butyl derivative 15, and 17:83 e.r. with 9-
anthracenyl catalyst 16, entries 11 and 12 respectively).
Quininium catalyst 17 gave better e.r. (88:12), and a switch
to a weaker base with solid K3PO4 gave similar e.r. (89:11;
entry 13). Changing to aqueous K3PO4 led to an increase in
e.r. (to 97:3) without the issues of ester hydrolysis that
plagued our earlier attempts to use aqueous base. We believe
this to be a rare example of a catalytic enantioselective C-
acylation of an enolate.[20]
Table 1: Optimization: cation-directed enantioselective C-acylation.[a]
With an optimized procedure in hand, we examined the
scope and limitations of this reaction (Table 2). The unsub-
stituted C2-symmetric spirobiindanone 9 can be produced in
93% yield and 97:3 e.r. via this approach; this cyclization can
be performed on a gram scale without compromising yield or
enantioselectivity. The absolute configuration of this material
was confirmed to be (S)- through comparison with literature
data.[21a] Substitution is tolerated on both the indan-1-one and
benzoic acid sections of substrates; thus C2-symmetric
dimethyl 18 (92:8 e.r., 84% yield) and bistrifluoromethyl
spiro derivatives 19 (95:5 e.r., 95% yield) are both produced
with excellent enantioselectivity and yield. This represents
a novel method for the catalytic enantioselective synthesis of
these axially chiral motifs; other methods rely on resolution
or the use of chiral auxiliaries.[21] The reaction is not limited to
C2-symmetric products and we have explored substitution at
all sp2 positions on the 1-indanone ring. Cyclization of
substrates bearing substituents in the 5-position of the
indanone ring occured with high levels of enantioselectivity
to afford 20 (5-bromo) and 21 (5-fluoro) with consistently
high levels of enantioselectivity (96:4 e.r. in both cases). The
absolute configuration of bromide 20 was confirmed through
X-ray crystallography.[22] An indanone substrate substituted
with an electron-donating substituent in the 4-position
cyclized to afford methoxy derivative 22 in 99% yield and
94:6 e.r. Similarly, cyclization of a 4-bromo indanone to give
23 occurred in 93:7 e.r. We have also examined the formation
of spirobiindanones with different groups on each of the two
Entry
R1
Catalyst
Base
e.r.[b]
1
2
3
4
5
6
7
8
Ph
Ph
Ph
Ph
Ph
Ph
C6F5
C6F5
C6F5
C6F5
C6F5
C6F5
C6F5
C6F5
C6F5
Bu4NBr
10
10
10
11
11
11
12
13
14
15
16
17
17
17
K2CO3 (aq.)[c]
K2CO3 (aq.)[c]
KOH (aq.)[c]
KOH (s)
KOH (s)
NaOPh (s)
KOH (s)
KOH (s)
KOH (s)
KOH (s)
KOH (s)
50:50
56:44
–
59:41
69:31
51:49
42:58
79:21
25:75
36:64
29:71
17:83
88:12
89:11
97:3
9
10
11
12
13
14
15
KOH (s)
KOH (s)
K3PO4 (s)
K3PO4 (aq.)[c]
[a] Conditions: substrate 7 or 8 (0.02 mmol), catalyst (10 mol%), solid
base (1.0 equiv.), PhMe ([substrate]=0.1 moldmÀ3), RT, 48 h. [b] e.r.
determined by chiral stationary phase HPLC. [c] base: 50% aq., w/w,
10.0 equiv.
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!