Dirhodium(II)-catalyzed intramolecular C-H amination of aryl azides

Article abstract of DOI:10.1002/anie.200800689

(Chemical Equation Presented) Azides to indoles: Dirhodium(II)-catalyzed decomposition of aryl azides was developed as a mild, functional group tolerant method for the synthesis of indoles (see scheme).

Full text of DOI:10.1002/anie.200800689

Communications
DOI: 10.1002/anie.200800689
Indole Synthesis
ꢀ
Dirhodium(II)-Catalyzed Intramolecular C H Amination of Aryl
Azides**
Meihua Shen, Brooke E. Leslie, and Tom G. Driver*
The development of new transition-metal-catalyzed methods
for selective functionalization of carbon–hydrogen bonds
continues to be an active area of research.[1] Whereas many
transition-metal complexes exhibit activity,[1] rhodium(II)
dimers[1c,2] are well established to react with a-diazo com-
pounds[3,4] or sulfonyliminoiodinanes[5,6] to access metal
carbenoids or nitrenoids,which can functionalize proximal
would enable the rhodium-catalyzed synthesis of indoles by
ꢀ
the formation of either the aryl C N bond (from azidoacry-
ꢀ
lates) or the creation of the vinyl C N bond (from aryl azides;
Scheme 1). As the thermal variant of this reaction,the
Sundberg indole synthesis,[8e] requires heating of the poten-
ꢀ
ꢀ
ꢀ
aliphatic C H bonds to form new C C bonds or C N bonds
in a stereoselective manner. Transition-metal-mediated for-
ꢀ
mation of new carbon–nitrogen bonds from vinyl or aryl C H
bonds,however,is much less common.
[7]
Azides can be employed in the amination of aromatic or
ꢀ
vinyl C H bonds. Thermolysis or photolysis of azides
produces nitrenes,[8,9] which react with proximal C H bonds
ꢀ
to form N-heterocycles.[10] Nitrenes,however,are highly
reactive and can decompose into a variety of byproducts,
[8b]
including amines,azobenzenes,or tars.
Whereas metal-
Scheme 1. Complimentary rhodium(II)-catalyzed indole syntheses.
mediated nitrogen atom transfer reactions from azides are
well-known to attenuate this extreme nitrene reactivity,[11,12]
dirhodium(II) carboxylates have been rarely employed to
catalyze these processes[13–15] despite their proven utility in
other related atom transfer reactions. Since azides are readily
tially explosive aryl azide and produces indole in variable
yields,transition-metal-catalysis of this reaction at a reduced
temperature would enhance the synthetic utility of this
transformation.
available,[10c,16,17] their use in new transition-metal-mediated
[18]
ꢀ
methods that create new C N bonds is highly appealing.
We recently discovered that indoles could be generated
from azidoacrylates 2 through exposure to catalytic amounts
of rhodium(II) perfluorobutyrate.[18a] Whereas this reaction
exhibited a broad substrate scope,it required an a-azidome-
thylacetate,which restricted product formation to 2-indole-
carboxylate esters.[17c] Achievement of indole synthesis from
Toward this goal,the potential for a transition-metal-
ꢀ
mediated vinyl C H bond amination reaction of 4 was
investigated (Table 1). We found that the use of rhodium(II)
catalysts with electron-deficient carboxylate ligands afforded
promising yields and conversions (Table 1,entries 2–4). The
use of 100 wt% of crushed 4 Š molecular sieves enabled the
reduction of the catalyst loading to 2 mol%,and was required
to produce reproducible results (Table 1,entry 3). Exposure
of the aryl azide to the recovered molecular sieves did not
result in any indole production. The lack of reactivity by the
recovered molecular sieves indicates that rhodium is not
being deposited onto the solid support to create a new
heterogeneous catalyst. Whereas our earlier study identified
rhodium(II) perfluorobutyrate as the only competent cata-
lyst,[18a] indole formation from aryl azide 4 could be mediated
as well by rhodium(II) octanoate (Table 1,entry 6). With this
rhodium complex,12,-dichloroethane was found to be the
superior solvent (Table 1,entry 7). Employing other less
soluble or more sterically congested rhodium(II) carboxylate
or rhodium(II) lactamate complexes provided only trace
indole formation (Table 1,entries 5,8,and 9). As we observed
in our previous method,no reaction was detected with other
metal salts[11] (such as Ag,[21] Cu,[6b] Au,[22] Co,[23] and Fe
complexes[24]) or Lewis (or Brønsted) acids[25] reported to
catalyze nitrenoid formation from azides or other nitrene
equivalents.[20]
ꢀ
aryl azides 3 through rhodium-catalyzed vinyl C H bond
amination would address this limitation as a broader range of
aryl azides are readily available from commercial starting
materials in two mild,functional group tolerant steps:
palladium-catalyzed Suzuki cross-coupling of 2-bromoani-
lines[19] and subsequent diazo transfer would produce 3.[20] The
combination of this potential method with our earlier one
[*] Dr. M. Shen, B. E. Leslie, Prof. Dr. T. G. Driver
Department of Chemistry
University of Illinois at Chicago
845 W. Taylor St., Chicago, IL 60607 (USA)
Fax: (+1)312-996-0431
E-mail: tgd@uic.edu
[**] This research was financially supported by the Petroleum Research
Fund administered by the American Chemical Society (46850-G1)
and the University of Illinois at Chicago. We thank Dr. Dan McElheny
(UIC) for assistance with NMR spectroscopy, and Dr. Furong Sun
(UIUC) for mass spectrometry data.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
5056
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5056 –5059

Angewandte
Chemie
The importance of the identity of the R5 substituent on the
Table 1: Optimization of catalytic conditions to form 2-phenylindole.
ꢀ
efficiency of the C H amination reaction prompted a more
thorough study of its nature (Table 3). As long as this
Table 3: Interrogation of the electronic influence of substituents on
indole formation.
Entry Metal salt
mol[%] 4 Š M.S. [wt%] Yield [%][a]
1
none
n.a.
5
0
0
0
63
Entry
6
R2
R3
R5
Yield [%][a]
2
3
4
5
[Rh2(O2CC3F7)4]
[Rh2(O2CC3F7)4]
[Rh2(O2CCF3)4]
[Rh2(OAc)4]
2
5
10
2
2
5
0
100
0
100
100
0
>95
33
trace
58
>95
trace
1
2
3
4
5
6
7
8
9
l
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OMe
F
OCF3
CF3
H
4-MeC6H4
4-FC6H4
3-FC6H4
4-ClC6H4
4-(F3C)C6H4
4-(F3C)C6H4
4-(F3C)C6H4
4-(F3C)C6H4
Me
96
95
91
84
95
90
89
91
m
n
o
p
q
r
s
t
u
6
[Rh2(O2CC7H15)4]
[Rh2(O2CC7H15)4]
[Rh2(O2CtBu)4]
7[b]
8
9
5
0
trace
61[b] (75)[c]
68[d] (66)[e]
10
n-C6H13
[a] As determined by 1H NMR spectroscopy by using CH2Br2 as an
internal standard. [b] 1,2-Dichloroethane was employed as the solvent.
M.S.=molecular sieves; n.a.=not applicable.
[a] Yield of product isolated after flash chromatography on SiO2. [b] 15%
of the azide recovered. [c] 5 mol% of [Rh2(O2CC7H15)4] employed in 1,2-
dichloroethane; 19% of the azide recovered. [d] Starting azide was a 98:2
mixture of E/Z isomers. [e] 5 mol% of [Rh2(O2CC7H15)4] employed in 1,2-
dichloroethane; 28% of the azide recovered.
With the realization of an optimized procedure,the scope
and limitations of this indole synthesis were investigated
(Table 2). The methodology was tolerant of a variety of
electronically different para substituents to the azide
(Table 2,entries 1–6). The importance of the molecular
sieve additive is illustrated in Table 2,entries 1 and 4; without
molecular sieves,the yield of the process dropped precip-
itously. Azides with a variety of different R1 and R2
substituents were also submitted to reaction conditions to
consistently provide 6- or 7-substituted indoles (Table 2,
entries 7–10). If the phenyl group was moved to the a position
(R4 group),however,the reactivity of the substrate was
reduced (Table 2,entry 11). Employing rhodium perfluoro-
butyrate as the catalyst produced only trace indole,whereas,
rhodium octanoate provided only 42% of the indole product.
substituent was an aryl group,modulation of its electronics
had little impact on the yield of the reaction (Table 3,
entries 1–5). Adjusting the electronic nature of the arene,
bearing the azide group,through R 2 or R3 substitution did not
affect the yield or conversion of the reaction either (Table 3,
entries 6–8). Changing the identity of the R5 substituent from
an aryl group to an alkyl group,however,reduced both the
conversion and yield (Table 3,entries 9 and 10). Rhodium
octanoate proved to be a more efficient catalyst for these
substrates,forming 75% indole versus 61% when rhodium
perfluorobutyrate was employed (Table 3,entry 9).
Whereas the exact mechanism remains unclear,several
different pathways could account for the indole formation
(Scheme 2). Rhodium could function as a nitrogen atom
transfer reagent.[2c,6a,26] In this mechanism,coordination of the
rhodium carboxylate to the a nitrogen[27] or the terminal
Table 2: Scope of rhodium-catalyzed indole formation.
Entry
6
R1
R2
R3
R4
R5
Yield [%][a]
1
2
3
4
5
6
7
8
9
a
b
c
d
e
f
g
h
i
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
OMe
Me
CF3
H
OMe
Me
H
H
H
H
H
H
H
H
H
H
H
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
98 (63)[b]
89
94
F
99 (75)[b]
OCF3
CF3
H
H
H
95[c]
82
88
89
89
10
11
j
k
F
H
72 (92)[e]
trace[d] (39)[e]
H
[a] Yield of isolated product after flash chromatography on SiO2. [b] No
M.S. [c] 2 mol% [Rh2(O2CC3F7)4]; 90% yield. [d] 67% recovered azide.
[e] 5 mol% of [Rh2(O2CC7H15)4] employed in 1,2-dichloroethane.
Scheme 2. Possible rhodium-catalyzed mechanisms to form 2-substi-
tuted indoles.
Angew. Chem. Int. Ed. 2008, 47, 5056 –5059
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5057

Communications
g nitrogen[28] of the azide (to form 10) induces loss of N2 to
Experimental Section
form nitrenoid 11 (or its resonance form 13).[29] Functional-
General procedure for the synthesis of 2-arylindoles: Toluene
(0.20 mL) or 1,2-dichloroethane (0.20 mL) was added to a mixture
of aryl azide (0.300 mmol),[Rh 2(O2CC3F7)4] or [Rh2(O2CC7F15)4]
(5 mol%),and crushed 4 Š molecular sieves (100 wt%) to a final
concentration of 1.5m. The resulting mixture was heated to 608C.
After 16 h the reaction mixture was cooled to room temperature and
the heterogenous mixture was filtered through SiO2. The filtrate was
concentrated in vacuo. Purification by using MPLC methods (EtOAc/
hexanes 0:100!30:70) provided the indole as a white solid.
ꢀ
ization of the proximal C H bond by 11 could occur in a
ꢀ
ꢀ
stepwise fashion with C N bond formation preceding C H
bond cleavage[30] through 12 or through a concerted transition
state (such as 14). Alternatively,C N bond formation could
occur simultaneously with N2 extrusion in a Schmidt-like
mechanism to form 12 directly from 10.[31] Subsequent to the
formation of 12,a 15,-hydride shift would produce the 1 H-
indole.[32]
ꢀ
Received: February 12,2008
Revised: April 4,2008
Published online: May 30,2008
ꢀ
ꢀ
If C N bond formation precedes C H bond cleavage,
then the Z isomer of 4 might be expected to form indole 5. To
test this hypothesis,( Z)-4 was submitted to reaction condi-
tions and found to afford 2-phenylindole [Eq (1)].[33] Higher
ꢀ
Keywords: azides · C H amination · indoles · rhodium
yields (93% versus 64%) were obtained when [Rh -
.
2
(O2CC7H14)4] was employed as the catalyst. Examination of
(Z)-4 during the course of the reaction revealed no E/
Z isomerization of the aryl azide.[34] The comparable yields of
indole formation from both (Z)- and (E)-4 differs from the
corresponding thermal reaction,which provided 2-phenyl-
indole from (Z)-4 in only 18% yield.[30a]
[1] For recent reviews,see: a) K. Godula,D. Sames, Science 2006,
312,67; b) A. R. Dick,M. S. Sanford, Tetrahedron 2006, 62,
2439; c) H. M. L. Davies, Angew. Chem. 2006, 118,6574; Angew.
Chem. Int. Ed. 2006, 45,6422.
[2] For recent reviews,see: a) P. Müller,C. Fruit, Chem. Rev. 2003,
103,2905; b) J. Du Bois, Chemtracts 2005, 18,1; c) M. P. Doyle, J.
Org. Chem. 2006, 71,9253; d) C. J. Moody, Angew. Chem. 2007,
119,9308; Angew. Chem. Int. Ed. 2007, 46,9148.
[3] T. Ye,M. A. McKervey, Chem. Rev. 1994, 94,1091.
[4] M. P. Doyle,M. A. McKervey,T. Ye, Modern Catalytic Methods
for Organic Synthesis with Diazo Compounds: From Cyclo-
propanes to Ylides,Wiley,New York, 1998.
[5] a) R. Breslow,S. H. Gellman, J. Am. Chem. Soc. 1983, 105,6728;
b) D. Mansuy,J.-P. Mahy,A. DurØault,G. Bedi,P. Battioni,
J.
Chem. Soc. Chem. Commun. 1984,1161.
A stepwise mechanism is additionally supported by the
reactivity of gem-diphenyl-substituted 15 [Eq. (2)]. Exposure
of 15 to the rhodium(II) carboxylate catalyst resulted in the
formation of 2,3-diphenyl indole 17. This 2,3-phenyl shift
implicates the intermediacy of cation 16. Together with the
reactivity of (Z)-4,these results suggest that indole formation
[6] Generated in situ: a) C. G. Espino,J. Du Bois, Angew. Chem.
2001, 113,618; Angew. Chem. Int. Ed. 2001, 40,598; b) P.
Dauban,L. Sani›re,A. Tarrade,R. H. Dodd, J. Am. Chem. Soc.
2001, 123,7707.
[7] Recent reports: a) X. Chen,X.-S. Hao,C. E. Goodhue,J.-Q. Yu,
J. Am. Chem. Soc. 2006, 128,6790; b) H.-Y. Thu,W.-Y. Yu,C.-M.
Che, J. Am. Chem. Soc. 2006, 128,9048; c) A. R. Dick,M. S.
Remy,J. W. Kampf,M. S. Sanford, Organometallics 2007, 26,
ꢀ
occurs by a mechanism where C N bond formation precedes
[35,36]
ꢀ
N H bond formation.
1365; d) Z. Li,D. A. Capretto,R. O. Rahaman,C. He,
Chem. Soc. 2007, 129,12058.
J. Am.
[8] a) P. A. S. Smith,B. B. Brown, J. Am. Chem. Soc. 1951, 73,2435;
b) P. A. S. Smith,J. H. Hall, J. Am. Chem. Soc. 1962, 84,480;
c) H. Hemetsberger,D. Knittel,H. Weidmann, Monatsh. Chem.
1970, 101,161; d) R. J. Sundberg,H. F. Russell,W. V. Ligon,Jr.,
L.-S. Lin, J. Org. Chem. 1972, 37,719.
[9] Azirines can also function as nitrene precursors. For recent work,
see: D. F. Taber,W. Tian, J. Am. Chem. Soc. 2006, 128,1058.
[10] Reviews: a) E. F. V. Scriven,K. Turnbull, Chem. Rev. 1988, 88,
297; b) C. J. Moody, Comprehensive Organic Synthesis, Vol. 7,
(Eds.: B. M. Trost,I. Fleming,S. Ley),Pergamon,Oxford, 1991,
p. 21; c) B. C. G. Sꢀderberg, Curr. Org. Chem. 2000, 4,727.
[11] Reviews: a) T. Katsuki, Chem. Lett. 2005, 34,1304; b) S. Cenini,
E. Gallo,A. Caselli,F. Ragaini,S. Fantauzzi,C. Piangiolino,
Coord. Chem. Rev. 2006, 250,1234.
[12] For recent reports,see: a) G.-Y. Gao,J. E. Jones,R. Vyas,J. D.
Harden,X. P. Zhang, J. Org. Chem. 2006, 71,6655; b) C.
Piangiolino,E. Gallo,A. Caselli,S. Fantauzzi,F. Ragaini,S.
Cenini, Eur. J. Org. Chem. 2007,743; c) J. V. Ruppel,R. M.
Kamble,X. P. Zhang, Org. Lett. 2007, 9,4889.
Dirhodium(II)-catalyzed decomposition of aryl azides
was developed as a mild,functional group tolerant method
for the synthesis of indoles. Future quantitative studies will
probe the different reaction trends observed with rhodium(II)
perfluorobutyrate and octanoate to ascertain any mechanistic
differences between the two catalysts. The conclusions of
these studies will be used to develop new transition-metal-
mediated methods that form N-heterocycles from azides.
[13] For a report where NsN3 was decomposed photochemically in
the presence of styrene and
a
rhodium(II) bis-
(naphthol)phosphate catalyst to produce an aziridine in 9%
5058
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5056 –5059

Angewandte
Chemie
yield,see: P. Mueller,C. Baud,I. Naegeli, J. Phys. Org. Chem.
1998, 11,597.
[26] For leading mechanistic discussions of RhII-mediated nitrogen
atom transfer from iminoiodinanes,see: a) K. W. Fiori,J.
Du Bois, J. Am. Chem. Soc. 2007, 129,562; b) C. Liang,F.
Collet,F. Robert-Peillard,P. Müller,R. H. Dodd,P. Dauban,
Am. Chem. Soc. 2008, 130,343.
[14] For the use of [Rh2(OAc)4] to accelerate the transformation of a
single vinyl azide to indole at 1158C,see: J. Sawyer,D. W.
Beight,E. C. R. Smith,D. W. Snyder,M. K. Chastain,R. L.
Tielking,L. W. Hartley,D. G. Carlson, J. Med. Chem. 2005, 48,
893.
J.
[27] For reports of Lewis acid coordination to the a-nitrogen atom of
an azide,see: a) ref. [25b]; b) G. A. Olah,P. Ramaiah,Q. Wang,
G. K. S. Prakash, J. Org. Chem. 1993, 58,6900; c) A. Wrobleski,
J. AubØ, J. Org. Chem. 2001, 66,886.
[15] For a related Rh2(O2CCF3)4-catalyzed isomerization of aryl
azirines,see: S. Chiba,G. Hattori,K. Narasaka,
Chem. Lett.
2007, 36,52.
[28] a) G. Proulx,R. G. Bergman, J. Am. Chem. Soc. 1995, 117,6382;
b) H. V. R. Dias,S. A. Polach,S.-K. Goh,E. F. Archibong,D. S.
Marynick, Inorg. Chem. 2000, 39,3894.
[16] For reviews of azide reactivity,see: a) G. LꢁabbØ, Angew. Chem.
1975, 87,831; Angew. Chem. Int. Ed. Engl. 1975, 14,775; b) S.
Brꢂse,C. Gil,K. Knepper,V. Zimmermann, Angew. Chem. 2005,
117,5320; Angew. Chem. Int. Ed. 2005, 44,5188.
[29] For leading mechanistic and computational studies of related
RhII-mediated carbene transfer reactions,respectively,see:
[17] Aryl and vinyl azides are easily synthesized. See: a) referen-
ce [8b]; b) reference [10a]; c) L. Henn,D. M. B. Hickey,C. J.
Moody,C. W. Rees, J. Chem. Soc. Perkin Trans. 1 1984,2189;
d) F. W. Fowler,A. Hassner,L. A. Levy, J. Am. Chem. Soc. 1967,
89,2077; e) Q. Liu,Y. Tor, Org. Lett. 2003, 5,2571; f) C.-Y. Liu,P.
Knochel, J. Org. Chem. 2007, 72,7106; g) E. D. Goddard-Borger,
R. V. Stick, Org. Lett. 2007, 9,3797.
a) M. C. Pirrung,H. Liu,A. T. Morehead,Jr.,
Soc. 2002, 124,1014; b) E. Nakamura,N. Yoshikai,M. Yama-
naka, J. Am. Chem. Soc. 2002, 124,7181.
J. Am. Chem.
ꢀ
ꢀ
N
[30] In the related thermal nitrene C H amination reaction,C
ꢀ
bond formation is believed to precede N H bond formation.
See: P. A. S. Smith,C. D. Rowe,D. W. Hansen,Jr., Tetrahedron
Lett. 1983, 24,5169.
[18] a) B. J. Stokes,H. Dong,B. E. Leslie,A. L. Pumphrey,T. G.
Driver, J. Am. Chem. Soc. 2007, 129,7500; b) H. Dong,M. Shen,
[31] For reviews of Lewis acid mediated Schmidt reactions involving
azide additions to olefins,see: a) W. R. Judd,C. E. Katz,J. AubØ,
Sci. Synth. 2005, 21,133; b) S. Lang,J. A. Murphy, Chem. Soc.
Rev. 2006, 35,146.
J. E. Redford,B. J. Stokes,A. L. Pumphrey,T. G. Driver,
Org.
Lett. 2007, 9,5191.
[19] For recent examples of anilines employed as substrates in the
Suzuki reaction,see: a) Y. Wu,J. Li,Y. Fu,Z. Bo, Org. Lett. 2004,
6,3485; b) Y. Miura,Y. Muranaka,Y. Teki, J. Org. Chem. 2006,
71,4786.
[20] Please refer to the Supporting Information for more details.
[21] Y. Cui,C. He, Angew. Chem. 2004, 116,4306; Angew. Chem. Int.
Ed. 2004, 43,4210.
[32] Alternatively,rhodium could function as a p Lewis acid to
activate the olefin towards attack by the pendant azide group.
For h2-olefin complexes involving [Rh2(O2CCF3)4],see: M. P.
Doyle,M. R. Colsman,M. S. Chinn, Inorg. Chem. 1984, 23,3684.
[33] Similar results were obtained with alkyl-substituted azides. See
the Supporting Information for more details.
[34] Neither [Rh2(O2CC3F7)4] nor [Rh2(O2CC7H15)4] were found to
isomerize cis-stilbene under the reaction conditions.
[22] Z. Li,D. A. Capretto,R. O. Rahaman,C. He, J. Am. Chem. Soc.
2007, 129,12058.
[35] Similar 1,2-phenyl shifts were observed in the thermolysis of
related b,b-disubstituted styryl azides (reference [30]) as well as
phosphine-mediated deoxygenation reactions of the correspond-
ing nitro compounds (R. J. Sundberg,T. Yamazaki, J. Org.
Chem. 1967, 32,290).
[23] F. Ragaini,A. Penoni,E. Gallo,S. Tollari,C. L. Gotti,M.
Lapadula,E. Mangioni,S. Cenini, Chem. Eur. J. 2003, 9,249.
[24] T. Bach,C. Kꢀrber, Tetrahedron Lett. 1998, 39,5015.
[25] a) H. Takeuchi,K. Takano, J. Chem. Soc. Perkin Trans. 1 1986,
611; b) H. Takeuchi,M. Maeda,M. Mitani,K. Koyama, J. Chem.
Soc. Perkin Trans. 1 1987,57.
[36] Attempts to trap cation 16 through the addition of allyltri-
methylsilane or allyltributylstannane were not successful.
Angew. Chem. Int. Ed. 2008, 47, 5056 –5059
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5059