Enantioselective friedel-crafts alkylations catalyzed by bis(oxazolinyl)pyridine-scandium(III) triflate complexes

Article abstract of DOI:10.1021/ja072976i

The enantioselective Friedel-Crafts addition of a variety of indoles catalyzed by bis(oxazolinyl)pyridine-scandium(III) triflate complexes (Sc(III)-pybox) was accomplished utilizing a series of β-substituted α,β-unsaturated phosphonates and α,β-unsaturate

Full text of DOI:10.1021/ja072976i

Published on Web 07/21/2007
Enantioselective Friedel-Crafts Alkylations Catalyzed by
Bis(oxazolinyl)pyridine-Scandium(III) Triflate Complexes
David A. Evans,* Keith R. Fandrick, Hyun-Ji Song, Karl A. Scheidt, and Risheng Xu
Contribution from the Department of Chemistry and Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138
Received April 27, 2007; E-mail: evans@chemistry.harvard.edu
Abstract: The enantioselective Friedel-Crafts addition of a variety of indoles catalyzed by bis(oxazolinyl)-
pyridine-scandium(III) triflate complexes (Sc(III)-pybox) was accomplished utilizing a series of â-substituted
R,â-unsaturated phosphonates and R,â-unsaturated 2-acyl imidazoles. The acyl phosphonate products
were efficiently transformed into esters and amides, whereas the acyl imidazole adducts were converted
to a broader spectrum of functionalities such as esters, amides, carboxylic acids, ketones, and aldehydes.
The sense of stereoinduction and level of enantioselectivity were found to be functions of the size of the
substrate employed, the substitution on the ligand, and the catalyst loading. Molecular modeling of the
catalyst with the bound substrates was performed based on the crystal structures of the catalyst complexes
and the sense of stereoinduction observed in the addition reaction. Nonlinear effects over a range of catalyst
concentrations implicate a mononuclear complex as the active catalyst.
Introduction
The Friedel-Crafts reaction and its enantioselective variants
are powerful carbon-carbon bond forming transformations in
organic chemistry.1 Due to the prevalence of the indole nucleus
in both natural products2 and potential medicinal agents,3
considerable effort has been extended in the development of
enantioselective alkylation reactions between this important
heterocycle and R,â-unsaturated carbonyl compounds.4 This
investigation documents our studies directed toward the devel-
opment of the chiral scandium(III)-pybox complex 1a for the
enantioselective alkylation of the indole and pyrrole nuclei, and
related electron-rich heteroaromatic substrates, with compatible
R,â-unsaturated carbonyl derivatives.
This investigation has focused on the evaluation of both
substrate structure, its chelating potential (eqs 1, 2), and catalyst
architecture for the development of an effective enantioselective
alkylation process.
Jørgensen was the first to extrapolate the Cu(II) chelation
motif in exploring this reaction utilizing our Cu(II)-bis-
(oxazoline) catalyst with R-keto esters and alkylidene malonates
(eqs 3, 4).5 Tang later improved on the use of arylidene
malonates in the Friedel-Crafts reaction by utilizing trisox-
azoline Cu(II) complexes.6 Concurrently, Umani-Ronchi re-
ported the addition reaction on two fronts: first, with the use
of R,â-unsaturated thioesters with a Pd(II)-(Tol-binap) catalyst,
and second, by designing a single-point binding catalyst for this
important transformation.7 During this present investigation,
Palomo demonstrated that R′-hydroxy enones are also competent
(1) For a review of the Friedel-Crafts reaction, see: Olah, G. A.; Krishnamurti,
R.; Prakash, G. K. S. Friedel-Crafts Alkylations. In ComprehensiVe
Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford,
1991; Vol. 3, pp 293-339.
(2) For examples of natural products which are relevant to the formed
stereocenter, see: cycloaplysinopsin (a) Mancini, I.; Guella, G.; Zibrowius,
H.; Pietra, F. Tetrahedron 2003, 59, 8757-8762. 10,11-Dimethoxynare-
line: (b) Kam, T.; Choo, Y. J. Nat. Prod. 2004, 67, 547-552. Hapalin-
doles: (c) Kinsman, A. C; Kerr, M. A. J. Am. Chem. Soc. 2003, 125,
14120-14125. (d) Huber, U.; Moore, R. E.; Patterson, G. M. L. J. Nat.
Prod. 1998, 61, 1304-1306.
(3) For examples of potential medicinal agents relevant to the formed
stereocenter, see: (a) Rawson, D. J.; Dack, K. N.; Dickinson, R. P.; James,
K. Bioorg. Med. Chem. Lett. 2002, 12, 125-128. (b) Dillard, R. D., et al.
J. Med. Chem. 1996, 39, 5119-5136. (c) Chang-Fong, J.; Rangisetty, J.
B.; Dukat, M.; Setola, V.; Raffay, T.; Roth, B.; Glennon, R. Bioorg. Med.
Chem. Lett. 2004, 14, 1961-1964.
(4) For a review of the asymmetric Friedel-Crafts reaction, see: Bandini, M.;
Melloni, A.; Umani-Ronchi, A. Angew. Chem., Int. Ed. 2004, 43, 550-
556.
(5) (a) Jensen, K. B.; Thorhauge, J.; Hazell, R. G.; Jørgensen, K. A. Angew.
Chem., Int. Ed. 2004, 40, 160-163. (b) Zhuang, W.; Hansen, T.; Jørgensen,
K. A. Chem. Commun. 2001, 347-348. For a review on the use of Cu-
(II)-bis(oxazoline) catalysts, see: (c) Johnson, J. S.; Evans, D. A. Acc.
Chem. Res. 2000, 33, 325-335.
(6) (a) Zhou, J.; Tang, Y. Chem. Commun. 2004, 432-433. (b) Zhou, J.; Ye,
M.-C.; Huang, Z.-Z.; Tang, Y. J. Org. Chem. 2004, 69, 1309-1320.
9
10.1021/ja072976i CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 10029-10041
10029

A R T I C L E S
Evans et al.
electrophiles for indole and pyrrole alkylations at ambient
temperatures with the same catalyst (eq 5).8 Organocatalytic
processes have also been recently reported (eq 7).9 Despite the
advances in the development of this reaction, there remains a
need for a general catalyst that functions under mild reaction
conditions.
8, 9)11 and those of Fukuzawa (eq 10)16 and Desimoni (eq 11).17
Subsequently, this catalyst complex has been used in enanti-
oselective Nazarov cyclizations18 and carbonyl ylide dipole
cycloadditions.19 In each of these examples, the substrates are
designed for chelate organization; accordingly, one might
conclude that chelate control is probably operating in these
reactions.
While Cu(II) complexes have proven to be effective chelating
chiral Lewis acids in this and related transformations (eqs
3-6),10 we have continued to expand the list of Lewis acids
that might be employed for these reactions. Our previous studies
identified bis(oxazolinyl)pyridine-scandium(III) (Sc(III)-py-
box) complexes as potential catalyst candidates for this
process.11-15 The present investigation presents a full account
of our studies in this area.
In addition to the illustrated allenylsilane additions (eqs 8,
9),11 we have also explored the utility of pybox-Sc(OTf)3
complexes in aldol additions to ethyl glyoxylate,12 in ene
reactions with ethyl glyoxylate and N-phenyl glyoxamide,13 and
in the addition of allylsilanes to N-phenylglyoxamide.14 This
report provides a full account of the development of the Friedel-
Crafts reaction utilizing scandium triflate complex 1a focusing
on reaction generality and the subsequent product elaborations
(Scheme 1).15 In order to access the full potential of this catalyst,
we explored the nature of the binding of R,â-unsaturated acyl
phosphonate and R,â-unsaturated 2-acyl imidazole substrates
to the catalyst with an emphasis on the geometry of the Sc-
(III)-PyBox catalyst-substrate complexes.
Chiral Sc(III) Complexes. The potential utility of pybox-
Sc(OTf)3 complexes as chelating chiral Lewis acids surfaced
in 2001 in independent publications from our laboratory (eqs
(7) (a) Bandini, M.; Melloni, A.; Tommasi, S.; Umani-Ronchi, A. HelV. Chem.
Acta 2003, 86, 3753-3763. (b) Bandini, M.; Fagioli, M.; Melchiorre, A.
M.; Umani-Ronchi, A. Tetrahedron Lett. 2003, 44, 5843-5846. (c) Bandini,
M.; Fagioli, M.; Garavelli, M.; Melloni, A.; Trigari, V.; Umani-Ronchi,
A. J. Org. Chem. 2004, 69, 7511-7518.
(8) Palomo, C.; Oiarbide, M.; Kardak, B. G.; Garc´ıa, J. M.; Linden, A. J. Am.
Chem. Soc. 2005, 127, 4154-4155.
(9) (a) Paras, N. A.; MacMillan, D. W. C. J. Am. Chem. Soc. 2001, 123, 4370-
4371. (b) Austin, J. F.; MacMillan, D. W. C. J. Am. Chem. Soc. 2002,
124, 1172-1173.
Results and Discussion
(10) For selected reviews of bis(oxazoline) Cu(II) catalysis, see: (a) Evans, D.
A.; Rovis, T.; Johnson, J. S. Pure Appl. Chem. 1999, 71, 1407-1415. (b)
See also ref 5c.
We initiated our investigation in the indole Friedel-Crafts
reaction with scandium(III)-pybox complex 1a by focusing on
(11) Evans, D. A.; Sweeney, Z. K.; Rovis, T.; Tedrow, J. S. J. Am. Chem. Soc.
2001, 123, 12095-12096.
(12) Evans, D. A.; Masse, C. E.; Wu, J. Org. Lett. 2002, 4, 3375-3378.
(13) Evans, D. A.; Wu, J. J. Am. Chem. Soc. 2005, 127, 8006-8007.
(14) Evans, D. A.; Aye, Y.; Wu, J. Org. Lett. 2006, 8, 2071-2073.
(15) Preliminary accounts of this work have been communicated: (a) Evans,
D. A.; Scheidt, K. A.; Fandrick, K. R.; Lam, H. W.; Wu, J. J. Am. Chem.
Soc. 2003, 125, 10780-10781. (b) Evans, D. A.; Fandrick, K. R.; Song,
H.-J. J. Am. Chem. Soc. 2005, 127, 8942-8943. (c) Evans, D. A.; Fandrick,
K. R. Org. Lett. 2006, 8, 2249-2252.
(16) Fukuzawa, S.; Matsuzawa, H.; Metoki, K. Synlett 2001, 5, 709-711.
(17) Desimoni, G.; Faita, G.; Filippone, S.; Mella, M.; Zampori, M. G.; Zema,
M. Tetrahedron 2001, 57, 10293-10212.
(18) (a) Liang, G.; Gradl, S. N.; Trauner, D. Org. Lett. 2003, 5, 4931-4934.
(b) Liang, G.; Trauner, D. J. Am. Chem. Soc. 2004, 126, 9544-9545.
(19) Suga, H.; Inoue, K.; Inoue, S.; Kakehi, A. J. Am. Chem. Soc. 2002, 124,
14836-14837.
9
10030 J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007

Friedel
−
Crafts Alkylations Catalyzed by Sc(III)
−pybox
A R T I C L E S
Scheme 1. Asymmetric Friedel-Crafts Alkylations Catalyzed by
Bis(oxazolinyl)pyridine-Scandium(III) Triflate Complex 1a
Table 1. Survey of Bidentate Enones in the Friedel-Crafts
Reaction with N-Methyl Indole Catalyzed by Complex 1aa
previously reported bidentate conjugate acceptors. Attempts at
the addition employing R,â-unsaturated acyl oxazolidinone 220
were disappointing as the reaction proceeded sluggishly (Table
1). We next considered the more reactive R,â-unsaturated
thiazolidine-2-thione 721 for the illustrated reaction. Instead of
the desired Michael adduct, we observed the formation of
bicyclic products 8a and 8b as racemates. We speculate that
the formation of product 8 arises from the initial intramolecular
attack of the thione to the unsaturated enone followed by the
Friedel-Crafts reaction with the nucleophilic indole.22 The R,â-
unsaturated acyl pyrazole 3 and the R,â-unsaturated Weinreb
amide 4 also provided trace amounts of product. The Friedel-
Crafts reaction with the previously employed aryl alkylidene
malonate 55,23 was sluggish and provided the desired product
in modest enantioselectivity (47% ee). Considering the previous
success our group has had with R,â-unsaturated acyl phospho-
nates in the hetero Diels-Alder reaction,24 we thought that these
substrates might provide the needed increase in reactivity. We
were pleased to see that the illustrated transformation with acyl
phosphonate 9a proceeded in quantitative yield and excellent
enantioselectivity (97% ee, Table 1). With the success of the
reaction with the acyl phosphonate 9a we next explored the
attributes of the addition reaction focusing on this substrate. It
should be noted that the intermediate â-indolyl acyl phospho-
nates were converted to the corresponding methyl esters (10a)
by direct addition of MeOH and DBU to the reaction mixture25
for the purpose of product characterization and analysis.
R,â-Unsaturated Acyl Phosphonate Substrates. With the
high stereoselectivity obtained in the indole alkylation utilizing
R,â-unsaturated acyl phosphonate 9a, we turned our attention
to optimizing ligand and reaction parameters. A pybox ligand
survey was conducted to determine the optimal ligand for the
illustrated transformation (Table 2). All of the pybox ligands
evaluated proved to be inferior to the Inda-pybox ligand 1a.
Interestingly, the Ph-pybox ligand 1b provided the opposite
sense of stereoinduction (-77% ee) as the structurally similar
a Reactions were performed at 0.1-0.4 M in substrate. b Enantiomeric
excess was determined by chiral HPLC (ODH, ADH, or OJH). c Initial
acyl phosphonate product was treated with MeOH and DBU to acquire the
corresponding methyl ester product 10a.
Table 2. Pybox Ligand Survey for the Reaction with Acyl
Phosphonate 9aa
catalyst
conversion (%)
% eeb
1a
1b
1c
1e
1f
99
99
99
99
99
98
-77
72
69
-77
(20) For some examples of the use of R,â-unsaturated acyl oxazolidinones from
our group, see: (a) Evans, D. A.; Scheidt, K. A.; Johnson, J. N.; Willis,
M. C. J. Am. Chem. Soc. 2001, 123, 4480-4491. (b) Evans, D. A.; Barnes,
D. M.; Johnson, J. S.; Lectka, T.; von Matt, P.; Miller, S. J.; Murry, J. A.;
Norcross, R. D.; Shaughnessy, E. A.; Campos, K. R. J. Am. Chem. Soc.
1999, 121, 7582-7594.
a All reactions were performed in CH2Cl2 at -78 °C and 0.13 M in
(21) Evans, D. A.; Miller, S. J.; Lectka, T.; von Matt, P. J. Am. Chem. Soc.
1999, 121, 7559-7573.
substrate. b Enantiomeric excess was determined by chiral HPLC analysis.
(22) A similar mechanism was proposed by Palomo in the sulfur transfer in
N-enoyl oxazolidine-2-thiones. Palomo, C.; Oiarbide, M.; Dias, F.; Ortiz,
A.; Linden, A. J. Am. Chem. Soc. 2001, 123, 5602-5603.
(23) For use of alkylidene malonates from our group, see: Evans, D. A.; Rovis,
T.; Kozlowski, M. C.; Downey, C. W.; Tedrow, J. S. J. Am. Chem. Soc.
2000, 122, 9134-9142.
Inda-pybox 1a complex (98% ee). A subsequent solvent survey
revealed that dichloromethane was optimal with regards to both
enantioselectivities and yields (Table 3). THF provided the
product in comparable enantioselectivity, but the reaction is less
efficient than when it is conducted with dichloromethane.
Diethyl ether and toluene were not effective solvents probably
(24) Evans, D. A.; Johnson, J. S.; Olhave, E. J. J. Am. Chem. Soc. 2000, 122,
1635-1649.
(25) Dialkyl acyl phosphonates are highly effective acylating agents. Sekine,
M.; Kume, A.; Hata, T. Tetrahedron Lett. 1981, 22, 3617-3620.
9
J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007 10031

A R T I C L E S
Evans et al.
Table 3. Solvent Survey for the Reaction with Acyl Phosphonate
Table 5. Friedel-Crafts Alkylations of Substituted Indoles with
R,â-Unsaturated 2-Acyl Phosphonate 9aa
9aa
solvent
time (h)
ee (%)b
yield (%)
CH2Cl2
THF
Et2O
4.3
40
40
97
96
NA
NA
94
50
1
indole
R1
R2
R3
R4
time (h)
ee (%)b
yield (%)
11b
11a
11c
11d
11e
11f
11g
11h
11i
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Br
Cl
3
21
5
83
96
98
99
83 (12b)
78 (10a)
76 (12c)
85 (12d)
toluene
46
3
Me
Allyl
Bn
Ts
Me
Me
Bn
Bn
Bn
Bn
Bn
a Reactions were performed at 0.13 M in substrate. b Enantiomeric excess
was determined by chiral HPLC analysis.
20
H
no reaction
Me
Ph
H
H
H
H
H
H
H
2
86
65
>99
>99
96
96
NA
99
94 (12f)
Table 4. Alkylations of N-Methylindole 11a with Representative
Acyl Phosphonates 9a
20
19
19
19
62 (12g)
64 (12h)
66 (12i)
67 (12j)
68 (12k)
trace
11j
11k
11l
OMe
CO2Me 17
NO2
Cl
CO2Me
H
H
H
6 d
20
47
11mc Bn
85 (12m)
68 (12n)
11nc
Bn
85
R
mol % 1a
temp (
°C)
time (h)
% eeb
yield (%)
Me (9a)
Me (9a)
Me (9a)
Me (9a)
Me (9a)
Et (9b)
10
10
10
5
-78
-56
-24
-78
-78
-50
-78
-78
-78
4
20
20
20
48
17
20
17
48
97
96
74
98
95
97
99
94
80
75 (10a)
99c (10a)
99c (10a)
88 (10a)
72 (10a)
65 (10b)
82 (10c)
57 (10d)
85 (10e)
a All reactions were performed at 0.2 M in substrate. b Enantiomeric
excess was determined by chiral HPLC analysis. c Reactions performed with
20 mol % 1a.
3
R,â-Unsaturated 2-Acyl Imidazole Substrates. Although
the acyl phosphonates 9 were found to be effective in the
Friedel-Crafts reaction with indoles there are drawbacks to this
substrate. First, the requisite R,â-unsaturated acyl phosphonates
9 were prepared in modest yields (20-47% yields), and the
parent acyl phosphonate itself is a sensitive “active ester”
functional group. Second, the reaction suffers from low substrate
tolerance (only â-alkyl substituted enones provided >90% ee),
and the reaction must be performed at e-50 °C to maintain
high enantioselectivities (>90% ee). In consideration of the
marginal stability of these active esters, we sought a new reactive
bidentate enone that would exhibit an increased level of stability
and comparable reactivity.
The 2-acyl imidazole group has been known as a latent acyl-
transfer reagent.26 Since R,â-unsaturated 1-acyl pyrazoles 3 have
been successfully used as bidentate conjugate acceptors in
asymmetric transformations,27 we felt that 2-acyl imidazoles 13
could be employed in the same capacity. We surmised that the
R,â-unsaturated 2-acyl imidazoles 13 would be more reactive
due to the decrease in electron donation into the carbonyl group
of the enone compared to the amide, ester, and imidate
counterparts. It was therefore not surprising that substrate 13a
performed well in the conjugate addition (eq 12).
10
10
10
20
i-Pr (9c)
CH2OTBDPS (9d)d
Ph (9e)
a All reactions were performed in CH2Cl2 at -78 °C (0.2 M). b Enan-
tiomeric excess was determined by chiral HPLC analysis. c Reported as
conversion based on 1H NMR spectroscopy. d Morpholine quench to acquire
the corresponding morpholine amide product: 94% ee, 73% yield.
due to catalyst insolubility. The effects of reaction temperature
were also evaluated, and it was demonstrated that the reaction
must be maintained e-50 °C to provide good enantioselectivity
(Table 4). In addition, the illustrated reaction could be performed
with as little as 3 mol % 1a without loss of enantioselectivity.
Our survey of the reaction variables reassured us that our initial
reaction parameters were optimal, and such were employed for
the rest of the study.
With the desired catalyst system in hand, the electrophile
scope was evaluated. The reaction is tolerant of alkyl â-substitu-
tion on the enone 9 (>95% ee, Table 4). However, the
corresponding reaction with the cinnamate derivative 9e requires
longer reaction times at higher catalyst loadings and displays a
moderate decrease in stereoinduction (80% ee).
We next turned our attention to the effects of substitution of
the indole ring in the illustrated transformation (Table 5). An
increase in the steric requirements of the nitrogen substituent
was shown to have a beneficial effect on stereoselectivity (indole
11b, 83% ee; N-methylindole 11a, 96% ee; N-benzylindole 11d,
99% ee, Table 5). However, an increase the steric requirements
at the 2-position of the indole nucleus leads to lower levels of
stereoinduction (1,2-dimethylindole 11f, 86% ee; 1-methyl
2-phenylindole 11g, 65% ee). Strong electron-withdrawing
groups on the indole nitrogen were shown to deactivate the
substrate (1-tosylindole 11e, and 4-nitroindole 11l). Besides
these limitations, the reaction is well tolerant of alkyl, methoxy,
and carboxyl-4 and -5 substitution on the indole ring providing
the products with excellent levels of selectivity (85-99% ee)
and in good yields (62-85% yield, Table 5).
Interestingly, the catalyst system with acyl imidazole 13a
produced the opposite enantiomer of the formed stereocenter
(26) For the transformation of 2-acyl benzimidazoles to esters, amides, â-dike-
tones, and â-ketoesters, see: (a) Miyashita, A.; Suzuki, Y.; Nagasaki, I.;
Ishiguro, C.; Iwamoto, K.-I.; Higashino, T. Chem. Pharm. Bull. 1997, 45,
1254-1258. (b) For the transformation of 2-acyl imidazoles to ketones,
â-diketones, â-ketoesters, and aldehydes, see: Ohta, S.; Hayakawa, S.;
Nishimura, K.; Okamoto, M. Chem. Pharm. Bull. 1987, 35, 1058-1069.
(27) Itoh, K.; Kanemasa, S. J. Am. Chem. Soc. 2002, 124, 13394-13395.
9
10032 J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007

Friedel
−
Crafts Alkylations Catalyzed by Sc(III)
−pybox
A R T I C L E S
Table 6. Solvent Survey for the Reaction with 2-Acyl Imidazole
13aa
solventb
temp (
°
C)
conv (%)c
time (h)
% eed
CH2Cl2
THF
toluene
CH3CN
CH2Cl2/THF
CH2Cl2/toluene
CH2Cl2/CH2CN
-78
99
17
17
99
81
89
99
5
48
48
17
48
48
48
75
79
33
89
30
60
79
-78 to -27
-78 to -27
-38
Figure 1. Catalyst loading profile for the Friedel-Crafts reaction with
R,â-unsaturated 2-acyl imidazole 13a (0.26 M in substrate, CH3CN, 4 Å
MS, -40 °C) and R,â-unsaturated acyl phosphonate 9a (0.13 M in substrate,
CH2Cl2, -78 °C) with N-methylindole (11a) catalyzed by Sc(III)-Inda-
pybox 1a.
-78
-78
-78 to -27
a Reactions were performed at 0.26 M in substrate. b Mixed solvent
systems refer to a 1:1 mixture of the corresponding solvents. c Conversion
based on 1H NMR spectroscopy. d Enantiomeric excess was determined by
chiral HPLC analysis.
Table 7. Effects of Water on the Friedel-Crafts Reaction with
R,â-Unsaturated 2-Acyl Imidazole 13aa
Figure 2. Temperature profile for the alkylation of N-methylindole (11a)
with 2-acyl imidazole 13a (1 mol % 1a, 0.26 M in substrate, CH3CN, 4 Å
MS) and acyl phosphonate 9a (10 mol % 1a, 0.13 M in substrate, CH3Cl2).
additive
conv. (%)b
% eec
none
99
99
99
80
87
90
84
53
15 mg 4 Å MS
0.05 equiv of H2O
0.5 equiv of H2O
Table 8. Ligand Survey for the Reaction between 2-Acyl
Imidazole 13a and N-Methylindole (11a)a
a Reactions were performed at 0.26 M in substrate. b Conversion based
on 1H NMR spectroscopy. c Enantiomeric excess was determined by HPLC
analysis.
as was seen in the corresponding reaction with acyl phospho-
nates 9a while proceeding with a decrease in enantioselection
(75% ee).
In order to improve the enantioselectivity of the addition
reaction, we turned our attention to the optimization of reaction
variables. A survey of solvents revealed acetonitrile as the choice
solvent, offering an increase in enantioselectivity to 89% ee
(Table 6). Water was found to be deleterious to the reaction
(Table 7), as the addition of 4 Å sieves to the reaction increased
the enantioselectivty of the transformation (90% ee) and afforded
more reproducible results. Conversely, the addition of 0.5 equiv
of water led to an erosion of selectivity (53% ee) and conversion
(80%) in the illustrated transformation.
Upon examination of the effects of catalyst loading, we
observed an inverse relationship between the amount of catalyst
1a employed and the enantioselectivity of the process (Figure
1) when R,â-unsaturated 2-acyl imidazole 13a was used as the
electrophile: 1 mol % (98% ee), 10 mol % (90% ee), 20 mol
% (78% ee), 50 mol % (11% ee). This rather dramatic trend
was most evident when a stoichiometric amount of catalyst 1a
was used, resulting in a turnover in asymmetric induction
(-31% ee). In contrast, when the reaction is performed with
acyl phosphonate 9a, we observed a rather flat catalyst loading
profile as usually seen with related catalytic systems: 1 mol %
(81% ee), 5 mol % (98% ee), 10 mol % (98% ee), 25 mol %
(95% ee), 100 mol % (93% ee).
catalyst
conversion (%)b
% eec
1a
1b
1c
1e
1g
91
78
55
39
19
98
93
96
92
36
a All reactions were performed at 0.26 M in substrate. bConversion based
on 1H NMR spectroscopy. cEnantiomeric excess was determined by chiral
HPLC analysis.
transformation with regards to both enantiofacial control and
reaction conversion. In contrast to the case with the acyl
phosphonates (Table 2), we did not observe a reversal in
selectivity when we employed the structurally similar Ph-pybox
catalyst 1b.
After determining the optimal ligand for the transformation,
the impact of reaction temperature was evaluated. The reaction
with imidazole 13a showed an excellent temperature profile
(Figure 2): the reaction could be performed at temperatures up
to 65 °C while maintaining excellent enantiofacial control (90%
We next explored the effects of reaction temperature (Figure
2) and ligand substitution (Table 8). As illustrated in Table 8,
the Inda-pybox complex 1a again proved to be optimal for the
9
J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007 10033

A R T I C L E S
Evans et al.
Table 9. Survey of R,â-Unsaturated 2-Acyl Heterocyclic Substrates in the Indole Friedel-Crafts Reaction Catalyzed by Sc(III)-Inda-pybox
Complex 1aa
a Reactions were performed at 0.26 M in substrate. b Enantiomeric excess was determined by chiral HPLC analysis.
Table 10. Survey of N-Substitution on the 2-Acyl Imidazole 13 in
the Indole Friedel-Crafts Reactiona
Table 11. Survey of â-Substitution on the R,â-Unsaturated 2-Acyl
Imidazole 13 in Reaction Catalyzed by Sc(III) Complex 1aa
% eeb
R
time (h)
yield (%)
% eeb
R
time (h)
yield (%)
Me (13a)
i-Pr (13b)
t-Bu (13c)
Ph (13d)
Bn (13e)
20
20
18
18
18
99
99
17
99
95
92 (14a)
95 (14b)
NA (14c)
91 (14d)
94 (14e)
Me (13a)
Et (13f)
n-Bu (13g)
i-Pr (13h)
CO2Et (13i)
Ph (13j)
3
12
8
12
12
8
93
97
95
78
95
94
71
86
93 (14a)
92 (14f)
93 (14g)
94 (14h)
96 (14i)
91 (14j)
91 (14k)
83 (14l)
a Reactions were performed at 0.26 M in substrate. b Enantiomeric excess
was determined by chiral HPLC analysis.
CH2OTBDPS (13k)
2-furan (13l)
8
8
a Reactions were performed at 0.26 M in substrate for 20 h. b Enantio-
meric excess was determined by chiral HPLC analysis.
ee). This is a notable contrast to the temperature profile observed
with acyl phosphonate 9a wherein low temperatures are required
for high levels of stereoinduction (e-50 °C, Figure 2).
Considering the nonbonding interactions in the binding pocket
of catalyst 1a we anticipated that the size of the imidazole
N-substituent could influence catalyst performance. As can be
seen in Table 10, an increase in the steric requirements of the
N-substituent on the substrate generally leads to an improvement
in reaction selectivity; however, this effect cannot be taken to
the extreme as the tert-butyl analogue 13c performed sluggishly
in the conjugate addition. Even though the isopropyl N-
substituted imidazole 13b was marginally better in overall yields
(99%) and selectivity (95%), we focused on the evaluation of
the more economical N-methylimidazoles 13a.
Attention was then directed toward an evaluation of the effects
of enone â-substitution (Table 11). The survey demonstrated
that the reaction is tolerant of alkyl, aryl, and even carboxylate
substitution at 0 °C.28 This is in contrast to observations with
acyl phosphonate 7, which only provided high selectivities
(>90% ee) at low temperatures (-78 °C) and only with â-alkyl
substituents.
Other heterocyclic R,â-unsaturated ketones were examined
to determine whether other candidate enone substrates might
also be viable for these indole alkylations (Table 9). When
benzimidazole 15 was employed, a modest decrease in enan-
tioselectivity was observed. We were pleased to see thiazole
17 performed well in the conjugate addition (97% yield, 94%
ee). Even though the thiazole 17 offered a marginal increase in
selectivity relative to the imidazole 13a, we decided to pursue
the imidazoles on the basis of practical considerations. In order
to ascertain the importance of the second nitrogen or sulfur atom
in the heterocycle, we evaluated the 2-pyridyl 19 and phenyl
21 substrates in the addition reaction. As suspected, these
substrates proved to display inferior selectivities (<32% ee).
We speculate that substrates 19 and 21 do not participate as
bidentate substrates for the reaction, and thus the second
heteroatom in the heterocycles (13, 15, and 17) must be vital
for chelate organization.
9
10034 J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007

Friedel
−
Crafts Alkylations Catalyzed by Sc(III)
−pybox
A R T I C L E S
Table 12. Alkylations of Indoles 11 with 2-Acyl Imidazole 13a
Catalyzed by Sc(III) Complex 1aa
Table 14. Survey of â-Substitution on the Imidazole 13 in the
Pyrrole Alkylationa
R
yield (%)
% eeb
Me (13b)
Et (13m)
i-Pr (13n)
CO2Et (13o)
Ph (13p)
4-MeOPh (13q)
4-MeO2CPh (13r)
2-furan (13s)
90
91
90
99
99
98
99
95
93 (17b)
86 (17m)
91 (17n)
84 (17o)
96 (17p)
92 (17q)
96 (17r)
91 (17s)
indole
R1
R2
R3
R4
mol % 1a
time (h)
ee (%)b
yield (%)
11b
H
H
H
H
H
Me
Ph
H
H
H
H
H
H
H
H
H
OMe
Br
Cl
Me
H
H
H
H
H
H
H
H
H
2.5
2.5
2.5
2.5
2.5
5
2.5
5
5
20
3
24
8
65
93
88
98
91
66
97
92
95
93
95
80 (15a)
93 (14a)
80 (15c)
90 (15d)
88 (15f)
43 (15g)
99 (15j)
55 (15h)
70 (15i)
91 (15o)
99 (15p)
11a Me
11c allyl
11d Bn
11f Me
11g Me
11j Bn
11h Bn
11i
11o Bn
11p Bn
2
90
20
20
20
20
20
a Reactions were performed at 0.13 M in substrate. b Enantiomeric excess
was determined by chiral HPLC analysis.
Bn
H
H
OMe
H
H
5
5
Table 15. Effects of Pyrrole N-Substitution with 2-Acyl Imidazole
13ba
a All reactions were performed at 0.26 M in substrate. b Enantiomeric
excess was determined by chiral HPLC analysis.
Table 13. Effects of N-Substitution of 2-Acyl Imidazole in the
Pyrrole Friedel-Crafts Reactiona
R
% eeb
yield (%)
H (16a)
Me (16b)
Bn (16c)
94
78
11
96 (17b)
89 (17t)
67 (17u)
a All reactions were performed at 0.13 M in substrate with 8 equiv of
pyrrole. b Enantiomeric excess was determined by chiral HPLC analysis.
pyrrole (8 equiv) were encouraging, as the reaction displayed
good levels of selectivity (87% ee, Table 13). From the modest
improvement in reaction selectivity in the indole Friedel-Crafts
reaction by increasing the steric demand of the N-substituent
on the 2-acyl imidazoles, we felt that this might also be in effect
for the corresponding pyrrole reactions. We were pleased to
see that this was true: as the N-i-Pr imidazole substrate 13b
provided the alkylation product in an improved enantiomeric
excess of 94%. In addition, the increased steric bulk of the
N-substituent on the imidazole had a beneficial effect on the
suppression of dimer formation (17-dimer, Table 13). Although
the phenyl N-substituted imidazole 13d was superior in overall
yield (98%) and selectivity (94% ee), we focused on the more
readily prepared N-i-Pr imidazole substrates with a small
sacrifice in yield for this study. As with the indole Friedel-
Crafts reaction with the 2-acyl imidazoles 13 (Figure 1), there
was an inverse relationship between catalyst loading (1a) and
enantioselectivity observed for the pyrrole counterpart. However,
due to the decreased concentration (0.13 M in substrate) used
for the pyrrole reactions, this effect was attenuated: 2 mol %
1a, 95% ee; 5 mol % 1a, 94% ee; 10 mol % 1a, 93% ee; 20
mol % 1a, 86% ee; 30 mol % 1a, 78% ee; 50 mol % 1a, 62% ee.
R
monomer/dimerb
yield (%)
% eec
Me (13a)
i-Pr (13b)
Ph (13d)
Bn (13e)
2.2:1d
>10:1
>20:1
>10:1
69
91
98
84
87 (17a)
94 (17b)
94 (17d)
91 (17e)
a Reactions were performed at 0.13 M in substrate for 20 h. b Monomer/
dimer ratio was determined by either 1H NMR spectroscopy or chiral HPLC.
c Enantiomeric excess was determined by chiral HPLC analysis. d Dimer
consisted of a 20:1 mixture of C2-symmetric (99% ee): meso-dimer (achiral,
not shown).
As was seen with acyl phosphonates 9 (Table 5), the reaction
with the R,â-unsaturated 2-acyl imidazoles 13 displays improved
enantiofacial control with N-benzyl substituted indoles (Table
12).15b Concurrent to that observed with the acyl phosphonates,
an increase in the steric requirements of the 2-substituent is not
well tolerated. However, slightly improved selectivity was seen
for 1,2-dimethylindole (91% ee) as compared to acyl phospho-
nate 9a (86% ee, Table 5). Halogen, alkyl, and methoxy 5- and
6-substituted N-benzylindoles were competent nucleophiles for
the addition reaction (92-97% ee, Table 12).
Pyrrole Alkylations. Due to our success in the indole
Friedel-Crafts reactions with R,â-unsaturated 2-acyl imidazoles
13, our attention was then directed toward the analogous pyrrole
alkylations. Initial attempts with 2-acyl imidazole 13a and
We were pleased to see that the pyrrole Friedel-Crafts
reaction can be conducted at 0 °C with minimal effects on
stereoinduction, and at this temperature, the reaction was
amenable to a wide range of â-substituted R,â-unsaturated 2-acyl
imidazoles 13 (Table 14). Alkyl-, aryl-, and carboxylate-
substituted enones were well tolerated (84-96% ee). Unlike
the examples in the indole study, the cinnamate-derived
(28) Improved selectivities were obtained when the reaction was performed at
-40 °C: R ) Me (1 mol% 1a) 97% yield, 97% ee; R ) Et (5 mol% 1a)
73% yield, 94% ee; R ) n-Bu (5 mol% 1a) 88% yield, 94% ee; R ) CO2-
Et (1 mol% 1a) 99% yield, 99% ee; R ) Ph (1 mol% 1a) 80% yield, 97%
ee; R ) CH2OTBDPS (2.5 mol% 1a) 83% yield, 96% ee; R ) 2-furan (5
mol% 1a) 99% yield, 86% ee.
9
J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007 10035

A R T I C L E S
Evans et al.
Table 16. Enantoselective Syntheses of 2-Substituted Indoles:
4,7-Dihydroindole Alkylation of 2-Acyl Imidazoles 13a
ethylaminoanisole 24 with R,â-unsaturated acyl phosphonate
9a proceeded cleanly to produce the methyl ester addition
product (with acyl phosphonate cleavage, methanol, and DBU)
in good yield and good selectivity (76% yield, 87% ee, eq 15).
However, the corresponding reaction with this nucleophile using
R,â-unsaturated 2-acyl imidazole 13a proceeded in poor yields
(<20%) and selectivity (22% ee). Contrary to that observed with
3-dimethylaminoanisole 24, the conjugate addition reaction with
2-methoxyfuran30 proved viable with the use of R,â-unsaturated
2-acyl imidazole 13a (65% yield, 98% ee, eq 16), yet the
corresponding addition reaction with R,â-unsaturated acyl
phosphonate 9a produced a complex mixture of products.
R
yield (%)
% eeb
Me (13b)
Et (13m)
i-Pr (13n)
Ph (13o)
4-MeOPh (13p)
4-CO2MePh (13q)
4-ClPh (13r)
2-ClPh (13s)
4-BrPh (13t)
2-furyl (13u)
99
97
62
98
97
85
98
90
85
92
95 (23b)
77 (23m)
72 (23n)
96 (23o)
90 (23p)
97 (23q)
96 (23r)
93 (23s)
95 (23t)
80 (23u)
a Reactions were performed at 0.13 M in substrate. b Enantiomeric excess
was determined by chiral HPLC analysis.
substrates are optimal (92-96% ee), whereas the fumarate 13o
was only moderately tolerated (84% ee).
While N-substitution in the indole alkylations improves
enantioselection, the opposite trend was noted with the analo-
gous pyrrole alkylations with optimal results being obtained with
no N-substituent on the heterocycle (Table 15). Concurrent to
these results, substitution adjacent to the reaction center (3-
position) is deleterious to reaction as can be seen in the
comparison illustrated in eqs 13 and 14.
Intramolecular Friedel-Crafts Alkylations. Although the
intermolecular Friedel-Crafts reaction has received attention,
the corresponding intramolecular examples are not common.31
We initially focused on the tethered N-benzyl indole 28b due
to the fact that N-benzyl indoles performed with the highest
stereoinduction in the intermolecular examples (Tables 5 and
12). However, this reaction was practically void of any
enantioselectivity (9% ee, Table 17). Considering the effects
of pyrrole N-alkyl substitution in the intermolecular reactions
(Table 15), we felt that an improvement in enantioselectivity
might be realized with an unalkylated indole substrate as the
ring is being formed at the 2-position of the indole nucleus in
the intramolecular reaction. Indeed, this was the case as the
reaction to form the six-membered ring with indole substrate
28c proceeded in excellent selectivity and yield (97% ee, 99%
yield). As seen with the pyrrole and indole examples, there was
an inverse relationship between enantioselectivity and mol %
1a employed (Table 17). We were surprised to observe that the
corresponding reactions to form the five- or seven-membered
rings were unsuccessful (28a and 28d).
With the success of the pyrrole alkylations, we felt that
dihydroindole (22, Table 16) would be a suitable nucleophile
for the conjugate addition, thus enabling the synthesis of
enantiomerically enriched 2-substituted indoles. This strategy
was first reported by Sarac¸oglu in the synthesis of 2-substituted
indoles via an oxidation of the intermediate dihydroindole
addition product.29 We were pleased to observe that the
corresponding reaction with 2-acyl imidazole 13b proceeded
well with this nucleophile (90% ee, 99% yield). The syntheses
of the enantiomerically enriched 2-substituted indole products
were accomplished in one-pot operations by the addition of
excess p-benzoquinone at the end of the conjugate addition
providing a range of 2-substituted indoles in good yields and
selectivities (Table 16).
Product Elaboration. Acyl phosphonates are well suited for
subsequent elaboration due to the inherent lability of the acyl
phosphoryl moiety. Thus, esters and amides are efficiently
accessed by treating the reaction product either with an alcohol
(30) For the use of 2-methoxyfuran in a conjugate addition reaction see: Itoh,
K.; Kitoh, K.; Sera, A. Heterocycles 1999, 51, 243-248.
3-Dimethylaminoanisole and 2-Methoxyfuran Alkylations.
Other aromatic nucleophiles were next evaluated in this
transformation. The Friedel-Crafts reaction utilizing 3-dim-
(31) (a) For nonasymmetric examples of intramolecular Friedel-Crafts alky-
lations, see: Bandini, M.; Melloni, A.; Tommasi, S.; Umani-Ronchi, A.
Synlett 2005, 8, 1199-1222. (b) For the asymmetric intramolecular Friedel-
Crafts alkylations with a tethered pyrrole to R,â-unsaturated carbonyl
substrates, see: Banwell, M. G.; Beck, D. A.; Smith, J. A. Org. Biomol.
Chem. 2004, 2, 157-159.
(29) C¸ avdar, H.; Sarac¸oglu, N. Tetrahedron 2005, 61, 2401-2405.
9
10036 J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007

Friedel
−
Crafts Alkylations Catalyzed by Sc(III)
−pybox
A R T I C L E S
Figure 3. X-ray structures of scandium(III) triflate-pybox(aquo) complexes 1a and 1b.
Figure 4. Coordination geometries for bidentate binding to Sc(III)-pybox complexes.
Table 17. Intramolecular Alkylations of Tethered Indolesa
esters, amides, â-diketones, and â-ketoesters are readily acquired
from the corresponding N-methylated 2-acyl imidazoles.26
Iodomethane alkylation in DMF at moderately elevated tem-
peratures proved to be successful as the methylated 2-acyl
imidazole salts were isolated in good yields. Excess alkylating
agent was removed with a stream of nitrogen, and the addition
of a variety of nucleophiles completed the one-pot cleavage
operation (Table 18). Esters (86-95% yield), amides (77-88%
yields), and the carboxylic acids (87%) were accessed using
this procedure. In addition, the 2-acyl imidazole 14a could be
transformed into a ketone or an aldehyde33-34 by initially treating
14a with a Grignard reagent to acquire the tertiary alcohol 32a
or alternatively reduced to the secondary alcohol 32b with
sodium borohydride (eqs 18 and 19).
imidazole
n
R
mol % 1a
temp (˚C)
yield (%)
% eeb
28a
28b
28c
28c
28c
28c
28d
0
1
1
1
1
1
2
H
Bn
H
H
H
H
H
5
5
2
0-rt
0
no conversion
99
99
9 (29b)
97 (29c)
96 (29c)
88 (29c)
79 (29c)
-40
-40
-40
-40
0-rt
5
99
20
50
10
99c
99c
decomposition
a Reactions were performed at 0.07 M in substrate. b Enantiomeric excess
was determined by chiral HPLC analysis. c Conversion based on 1H NMR
spectroscopy.
and an amine base (DBU) or with morpholine under mild
conditions (eq 17).
The resulting imidazole group was then methylated with
iodomethane in ethyl acetate and eliminated under basic
conditions to reveal the ketone 33 (87%) or aldehyde 34 (78%),
respectively.
Related transformations are illustrated in eqs 20-22. In the
first instance, the cleavage and cyclization of 2-acyl imidazole
On the other hand, the imidazoyl moiety of the 2-acyl
imidazoles may be transformed into an excellent leaving group
upon N-alkylation.26 The 2-H acidities of these imidazolium salts
are in the range 19-24 (pKa, DMSO) making them comparable
in acidity to an oxazolidinone.32 As such, aldehydes, ketones,
9
J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007 10037

A R T I C L E S
Evans et al.
17b to form the 2,3-dihydro-1H-pyrrolizine 35 was achieved
(eq 20). Methylation of 17b with methyl triflate (1.1 equiv,
MeCN, 25 °C, 2 h) was followed by the addition of a base (3
equiv DMAP or Hu¨nig’s base) to initiate the intramolecular
acylation in quantitative yield.
Figure 5. Crystal structures of catalyst complex 1a and 1b and stereo-
chemical models for complexes 1a and 1b with a bound R,â-unsaturated
acyl phosphonate substrate 9a (triflates and water molecules omitted for
clarity).
The cleavage of 2-acyl imidazoles 36 and 23o (eqs 21, 22)
may be achieved under the same conditions. Again, methyl
triflate (1.1 equiv, MeCN, 25 °C, 2 h) proved to be the most
reliable alkylation reagent. The subsequent addition of MeOH
and DBU to the reaction vessel provided the methyl ester 37 in
92% yield in a one-pot procedure. Indole 23o may also be
transformed into its methyl ester 38 in quantitative yield or the
corresponding carboxylic acid 39 in good yield. It should be
noted that the addition of 4 Å molecular sieves to scavenge
any residual moisture is necessary for reproducible results.
Figure 6. Eyring plot for the indole Friedel-Crafts reaction between R,â-
unsaturated acyl phosphonate 9a and Sc(III) complex 1a.
Table 18. Cleavage of 2-Acyl Imidazole 14a
Nuc(
−) conditions
Nuc
time
yield (%)
Mechanistic Studies
MeOH/DBU
EtOH/DBU
i-PrOH/DBU
H2O/DBU
i-PrNH2
-OMe
30 min
30 min
30 min
30 min
20 min
1 h
93 (31a)
86 (31b)
95 (31c)
87 (31d)
77 (31e)
88 (31f)
84 (31g)
-OEt
Crystal Structures. Although we have previously presented
octahedral models for the scandium(III)-pybox complex to
account for the sense of stereoinduction in addition reactions
involving glyoxylates11 and R,â-unsaturated acyl phosphonates,15a
we have been able to obtain crystal structures of the hydrates
-OCH(CH3)2
-OH
-NHCH(CH3)2
morpholine
-NHPh
morpholine
aniline
12 h
of both the Ph-pybox complex 1b11 and Inda-pybox complex
1a15a (Figure 3) that show a seven-coordinate pentagonal
bipyramidal geometry. We utilized this fact by employing the
seven-coordinate cationic Sc(III)-pybox (ScCl2(SbF6)-PyBox)
complexes with bidentate glyoxylates for asymmetric enolsilane
additons.12 However, with four available coordination sites on
the Sc(III)-pybox complexes, there is an ambiguity concerning
the binding of bidentate substrates to these complexes (Figure
4). We speculate that coordination complexes C-E might be
(32) (a) For the pKa in DMSO of the 2-proton of the N,N-dimethyl benzimi-
dazolium ion, see: Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456-463.
(b) For the pKa in DMSO of the 2-proton of the N,N-dimethyl imidazolium
ion, see: Alder, R. W.; Allen, P. R.; Williams, S. J. J. Chem. Soc., Chem.
Commun. 1995, 1267-1268. (c) For the pKa in water of the 2-protons
of the N,N-dialkyl imidazolium ions, see: Amyes, T. L.; Diver, S. T.;
Richard, J. P.; Rivas, F. M.; Toth, K. J. Am. Chem. Soc. 2004, 126, 4336-
4374.
(33) For the methylation and cleavage of â-hydroxyl 2-acyl imidazoles, see:
Davies, D. H.; Hall, J.; Smith, E. H. J. Chem. Soc., Perkin Trans. 1991, 1,
2691-2698.
(34) Davies, H. W.; Matasi, J. J.; Ahmed, G. J. Org. Chem. 1996, 61, 2305-
2313.
9
10038 J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007

Friedel
−
Crafts Alkylations Catalyzed by Sc(III)
−pybox
A R T I C L E S
Figure 7. Effects of catalyst concentration on enantioselectivity in the
indole Friedel-Crafts reaction of R,â-unsaturated 2-acyl imidazole 13a
catalyzed by Sc(III) complex 1a.
Figure 9. Eyring plot for the indole Friedel-Crafts reaction of R,â-
unsaturated 2-acyl imidazole 13a catalyzed by Sc(III) complex 1a.
Figure 10. Effects of conversion on the enantioselectivity in the indole
Friedel-Crafts reaction of R,â-unsaturated 2-acyl imidazole 13a catalyzed
by Sc(III) complex 1a (reaction performed with 50 mol % 1a at 0.026 M
in substrate at -40 °C).
extra ligand to occupy the equatorial plane.35 An examination
of the crystal structures of catalyst 1b (Sc(III)-Ph-pybox) and
1a (Sc(III)-Inda-pybox) shows the available spaces in the
equatorial planes of the two catalyst complexes are markedly
different (Figure 5). The “slot” in the equatorial plane of the
Ph-pybox complex 1b is 7.8 Å, whereas the corresponding space
in the Inda-pybox complex 1a is only 3.8 Å. Thus, the equatorial
position of the Sc(III)-Ph-pybox complex 1b can accommodate
the sterically demanding dimethyl phosphonate moiety of 9a.
This orientation positions the enone at the apical position, thus
exposing the Re-face of the enone to nucleophilic attack.
However, for the case of the Sc(III)-Inda-pybox complex 1a,
the space in the equatorial plane is reduced due to the inward
rotation of the phenyl rings. This, in effect, excludes the
sterically demanding dimethyl phosphonate from binding in the
equatorial position and positions it in the apical site. With the
enone bound at the equatorial position, the Si-face of the olefin
is available for nucleophilic attack (Figure 5). These models
are supported by the sense of stereoinduction observed in the
corresponding reactions with these complexes. Although the
apical position in complex 1b is the most sterically accessible
for the bulky phosphonate of 9a, there appears to be a preference
for the phosphonate residue in 9a to occupy the equatorial
position that overrides the inherent nonbonding interactions in
this position.
Figure 8. Nonlinear effects for the addition reaction of R,â-unsaturated
2-acyl imidazole 13a catalyzed by Sc(III) complex 1a: (a) 10 mol % 1a,
(b) 20 mol % 1a, (c) 40 mol % 1a.
sterically disfavored in preference to the octahedral complex
A. However, if smaller substrates are employed one might
imagine that complexes C-E could contribute to the addition
reactions.
Acyl Phosphonate 9a. The indole addition reaction of R,â-
unsaturated acyl phosphonate 9a catalyzed by scandium com-
plexes 1a (Inda-pybox) displays the opposite sense of stereoin-
duction to that observed in the reaction with the structurally
similar complex 1b (Ph-pybox). Considering the steric require-
ments of the bulky acyl phosphonate moiety, we speculate the
complexes are octahedral with the bound substrate 9a as
depicted in Figure 5 due to the limited space available for an
In further support of the stereochemical model, an Eyring
plot (Figure 6) of the addition reaction with R,â-unsaturated
(35) Modeling was performed by locking the ligand, metal, and one apical ligand
in place in the crystal structure utilizing Spartan ’04 (Windows XP, 3.0
GHz, 1.0 GB). The acyl phosphonate was then docked onto the apical
position and equatorial position (while locking the Sc-O distance to be
the same as to what is observed in the crystal structure for the corresponding
Sc-OTf bonds). A single-point energy minimization was performed
utilizing MMFF subjected to the constraints as stated above.
9
J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007 10039

A R T I C L E S
Evans et al.
Figure 11. Cylindrical-ball and space-filling models of catalyst complex 1b with bound R,â-unsaturated 2-acyl imidazole 13a and product molecule 14a.
acyl phosphonate 9a and catalyst 1a shows a linear relationship
between ln (% major enantiomer/% minor enantiomer) and the
reciprocal of the temperature (× 10-3 K-1), indicating that only
a single set of diastereomeric transition states are operating in
the reaction.36 We can interpret this result as indicating that a
single binding geometry is operating in the illustrated addition
reaction.
Acyl Imidazole 13a. The addition reaction with R,â-
unsaturated 2-acyl imidazoles 13a and Sc(III)-Inda-pybox
complex 1a shows the opposite sense of stereoinduction as
compared to the corresponding reaction with the acyl phos-
phonate 9a. In addition, this reaction displays an inverse
correlation between catalyst loading (mol % 1a) and enanti-
oselectivity (Figure 1), which is absent in the corresponding
reaction with the acyl phosphonate 9a. In order to ascertain
the cause of the catalyst loading profile we explored the effects
of catalyst concentration on enantioselectivity. As can be seen
in Figure 7, diluting the reaction does lead to a small im-
provement in enantioselectivity; however, this effect is more
evident when the catalyst concentration is decreased by lower-
ing the catalyst loading. This suggests that the observed
catalyst loading profile is not directly related to the catalyst
concentration.
Figure 12. Effects of enantioselectivity on the indole addition reaction of
To gain further insight into the nature of the catalyst-
substrate complex, we explored the effects of catalyst enantio-
meric excess versus reaction enantioselectivity (Figure 8). We
observed a linear relationship between the ligand enantiomeric
excess and the enantioselectivity of the addition reaction when
the analysis was performed with 10 mol % 1a. Because different
forms of the active catalyst complex could be more prevalent
at higher catalyst loadings, we conducted the same analysis at
20 and 40 mol % 1a. As seen in Figure 8, there are linear
relationships between catalyst ee and the enantioselectivity of
the addition reaction at these higher catalyst loadings as well.
Based on these results, we conclude the active catalyst is
mononuclear.
The Eyring plot (Figure 9) of the addition reaction with R,â-
unsaturated 2-acyl imidazoles 9a and catalyst 1a displays a linear
relationship between ln (% major enantiomer/% minor enanti-
omer) and the reciprocal of the temperature (× 10-3 K-1),
indicating that only a single set of diastereomeric transition states
are operating in this temperature range.36 We speculate that this
R,â-unsaturated 2-acyl imidazole 13a with (a) 3 equiv of propionyl 2-acyl
imidazole 40 and (b) 1 equiv of product 14a.
result is due to a single set of coordination geometries operating
in the addition reaction.
We next turned our attention to the effect of reaction
conversion on the enantioselectivity of the addition reaction.
This experiment was conducted with 50 mol % 1a to augment
this effect and at one-tenth the concentration in order to decrease
the reaction rate so that samples could conveniently be collected
at various reaction conversions. As seen in Figure 10, there is
an increase in enantioselectivity with higher reaction conversion,
thus indicating a possible beneficial effect of product formation
in the illustrated transformation.
Considering that the crystal structure of the hydrate of catalyst
complex 1a shows a pentagonal bipyramidal geometry and the
fact that the reaction displays an increase in enantioselectivities
with reaction conversion, we speculate that the reaction proceeds
through a seven-coordinate 1:1:1 product/substrate/catalyst
complex (Figure 11)37 that is favored at lower catalyst loading
and is more enantioselective than the corresponding 1:1
substrate/catalyst complex that would be favored at higher
(36) Buschmann, H.; Scharf, H.-D.; Hoffman, N.; Esser, P. Angew. Chem., Int.
Ed. Engl. 1991, 30, 477-515.
9
10040 J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007

Friedel
−
Crafts Alkylations Catalyzed by Sc(III)
−pybox
A R T I C L E S
catalyst loadings. The imidazole group is planar and less
sterically demanding than the dimethyl phosphonate moiety, and
thus two of these units can occupy the equatorial positions of
the catalyst, with one being the substrate and the other being
the product molecule. This would place the carbonyl groups at
the apical positions with the enone of the substrate oriented for
nucleophilic attack on the Re-face, in accord with a sense of
stereoinduction observed in the reaction. Inversion of the binding
of the substrate would be excluded as this would lead to a severe
steric interaction between the enone and the imidazole ring of
the bound product. We cannot help but notice the proximity of
the indole ring of the product to the pyridine ring of the bound
pybox ligand. Considering the electron-rich nature of the indole
ring and the electron deficient nature of the metal-bound pyridine
of the pybox ligand, we speculate there might be a beneficial
electrostatic interaction between these heterocycles that con-
tributes to the binding of a product molecule versus another
substrate in the seven-coordinate model.
reaction with 1 equiv of product and no ligand with 10 mol %
Sc(OTf)3 produced no noticeable enantioselectivity.
Conclusions
A highly enantioselective Michael-type indole Friedel-Crafts
reaction with a variety of â-substituted R,â-unsaturated acyl
phosphonates and â-substituted R,â-unsaturated 2-acyl imida-
zoles catalyzed by bis(oxazolinyl)pyridine-scandium(III) triflate
complexes have been developed. The addition reaction was
found to be more versatile with the R,â-unsaturated 2-acyl
imidazoles as pyrrole, indole, and intramolecular Friedel-Crafts
reactions were found to be highly stereoselective. The acyl
phosphonate products were efficiently transformed into esters
and amides, whereas the acyl imidazole products were converted
to more diverse functionalities such as esters, amides, a
carboxylic acid, a ketone, and an aldehyde. We also developed
a mild and efficient cleavage protocol for the diversification of
the 2-acyl imidazole products utilizing a slight excess of MeOTf
in acetonitrile as the key methylating condition. The binding
coordination of the bis(oxazolinyl)pyridine-scandium(III) tri-
flate complexes was probed by the series of experiments
performed with both the R,â-unsaturated acyl phosphonate and
the R,â-unsaturated 2-acyl imidazoles. The complexes appear
to adopt coordination geometries that are sterically favorable,
with the more bulky bidentate acyl phosphonate substrates
invoking octahedral geometries. Whereas, the less sterically
demanding 2-acyl imidazole substrates were able to adopt the
seven-coordinate pentagonal bipyramidal coordination geometry
seen in the crystal structures of the hydrate of scandium(III)
Ph-pybox 1b and Inda-pybox 1a complexes by invoking an
active catalyst that has a bound product molecule.
To further validate this hypothesis a series of experiments
were conducted where the catalyst loading of the reaction was
varied but included in the reaction a series of 2-acyl imidazole
additives, first with 3 equiv of the achiral saturated 2-acyl
imidazole 40 and second with 1 equiv of product 14a. (Figure
12).38 For the cases of low catalyst loading (<10 mol % 1a),
we did not notice a significant improvement on reaction facial
selectivities compared to the original catalyst loading profile.
However, when we raised the catalyst loading from 20 to 50
mol % 1a we observed a significant improvement in enanti-
oselectivity when the additives were present. To disprove any
stereoinduction from the product 14a, a control experiment was
performed with no catalyst and 1 equiv of product, and no
reaction was detected in this experiment. Also, the corresponding
Acknowledgment. Support is provided by the NIH (GM-
33328-20), the NSF (CHE-9907094), Merck, Amgen, and NSC
Technologies. H.-J.S. acknowledges a Novartis fellowship.
K.A.S. acknowledges the receipt of an NIH postdoctoral
fellowship.
(37) Modeling performed by locking the positions of the ligand, metal, triflates,
and water molecule in the crystal structure utilizing Spartan ’04 (Windows
XP, 3.0 GHz, 1.0 GB). The 2-acyl imidazole was then docked onto
the apical position and one of the available equatorial positions. A
product molecule was then docked onto one of the apical positions
and the remaining equatorial position. A single-point energy minimiza-
tion was performed utilizing MMFF subjected to the constraints as stated
above.
(38) Reactions were performed with 92% ee product as an additive (1 equiv)
and were allowed to proceed to 100% conversion. Conversion and
enantioselectivity were determined by chiral HPLC analysis. Enantiose-
lectivity of the reaction was determined by: eerxn ) 2eeobsd - 92.
Supporting Information Available: Experimental procedures,
spectra data for all new compounds, and stereochemical proof
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA072976I
9
J. AM. CHEM. SOC. VOL. 129, NO. 32, 2007 10041