Tertiary carbinamine synthesis by rhodium-catalyzed [3+2] annulation of N-Unsubstituted aromatic ketimines and alkynes

Article abstract of DOI:10.1002/chem.200902814

A convenient and waste-free synthesis of indene-based tertiary carbinamines by rhodium-catalyzed imine/ alkyne [3+2] annulation is described. Under the optimized conditions of 0.5-2.5 mol% [{(cod)Rh(OH)}₂] (cod=1,5- cyclooctadiene) catalyst 1,3-bis(diphenylphosphanyl) propane (DPPP) ligand, in toluene at 120 °C, N-unsubstituted aromatic ketimines and internal alkynes were coupled in a 1:1 ratio to form tertiary 1H-inden-1-amines in good yields and with high selectivities over isoquinoline products. A plausible catalytic cycle involves sequential imine-directed aromatic C-H bond activation, alkyne insertion, and a rare example of intramolecular ketimine insertion into a Rh^I-alkenyl linkage.

Full text of DOI:10.1002/chem.200902814

FULL PAPER
DOI: 10.1002/chem.200902814
Tertiary Carbinamine Synthesis by Rhodium-Catalyzed [3+2] Annulation of
N-Unsubstituted Aromatic Ketimines and Alkynes
Zhong-Ming Sun, Shuo-Ping Chen, and Pinjing Zhao*[a]
Abstract: A convenient and waste-free
synthesis of indene-based tertiary car-
binamines by rhodium-catalyzed imine/
alkyne [3+2] annulation is described.
Under the optimized conditions of 0.5–
2.5 mol% [{(cod)Rh(OH)}2] (cod=1,5-
cyclooctadiene) catalyst, 1,3-bis(diphe-
nylphosphanyl)propane (DPPP) ligand,
in toluene at 1208C, N-unsubstituted
aromatic ketimines and internal al-
kynes were coupled in a 1:1 ratio to
form tertiary 1H-inden-1-amines in
good yields and with high selectivities
over isoquinoline products. A plausible
catalytic cycle involves sequential
ꢀ
imine-directed aromatic C H bond ac-
Keywords: alkynes · carbinamines ·
tivation, alkyne insertion, and a rare
example of intramolecular ketimine in-
sertion into a RhI–alkenyl linkage.
ꢀ
C H activation
rhodium
·
ketimines
·
Introduction
Transition-metal-catalyzed tandem reaction sequences have
become a powerful strategy for the rapid assembly of com-
plex structures with simple substrates.[1] Over the past
ꢀ
decade, catalytic activation of C H bonds has increasingly
been exploited in various tandem reaction sequences.[2] This
approach towards reactive organometallic intermediates
ꢀ
through C H bond activation provides an eco-friendly alter-
native to conventional methods, for example, transmetala-
tion with main-group organometallic reagents, or oxidative
addition with organic halides and pseudohalides. Herein, we
Scheme 1. A proposed catalytic cycle for the transition-metal-catalyzed
[3+2] annulation of aromatic ketimines and alkynes.
ꢀ
ꢀ
report a catalytic C H bond activation/C C bond formation
tandem process for the construction of indene-based tertiary
carbinamines by a 1:1 coupling of aromatic ketimines with
alkynes.
metal–aryl linkage would generate an imine-chelated metal–
alkenyl intermediate B. Ring closure by an intramolecular
ketimine insertion into the metal–alkenyl linkage would
form an indene-based amido complex C.[3b] Subsequent
proton exchange would release the tert-carbinamine product
3 and regenerate the catalyst, which reacts with the ketimine
substrate to complete the catalytic cycle. This formal [3+2]
annulation would provide a convenient and waste-free com-
plement to current synthetic routes towards tertiary carbina-
mines.[4]
A key step in our design is the generation of a nucleophil-
ic transition-metal–alkenyl intermediate B that is reactive
towards intramolecular ketimine addition. Analogous late-
transition-metal–alkenyl and aryl nucleophiles have been
successfully accessed in catalytic tandem sequences initiated
by other types of elementary reactions.[5] For example, the
groups of Yamamoto and Lu have reported Pd-catalyzed in-
We envision a catalytic cycle that is initiated by the well-
ꢀ
documented heteroatom-directed C H bond activation
(Scheme 1).[2] In particular, imine-directed ortho-C H bond
ꢀ
activation with aromatic ketimine 1 would form a cyclome-
talated aryl complex A.[3a] Insertion of alkyne 2 into the
[a] Dr. Z.-M. Sun, S.-P. Chen, Prof. Dr. P. Zhao
Department of Chemistry and Molecular Biology
North Dakota State University
1231 Albrecht Avenue, P.O. Box 6050
Fargo, ND 58108-6050 (USA)
Fax : (+1)701-231-8831
E-mail: pinjing.zhao@ndsu.edu
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902814.
Chem. Eur. J. 2010, 16, 2619 – 2627
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2619

tramolecular ketone arylations initiated by oxidative addi-
tions with aryl bromides or transmetalation with arylboronic
acids.[5a–d] Very recently, Lautens and co-workers have dem-
onstrated the control over the electrophilic/nucleophilic dual
reactivity of arylpalladium species generated by oxidative
additions with aryl halides.[6] When such tandem sequences
late metal iminyl species are known to undergo a number of
elementary organometallic reactions that could compete
against the desired [3+2] annulations.[14–16]
Results and Discussion
ꢀ
are initiated by C H bond activations, however, formation
of nucleophilic organometallic intermediates appears to be
We began our investigation by studying catalytic reactions
between benzophenone imine (1a) and diphenylacetylene
(2a) (Table 1). Special attention was paid to rhodium-based
much more challenging.[7] In particular, aromatic ketones
ꢀ
and imines are known to undergo catalytic tandem C H
ꢀ
bond activation and addition with alkynes (or olefins) with-
out formation of carbocycles through intramolecular nucleo-
philic additions.[2,8] In contrast, the formation of heterocycles
catalysts due to their successful applications in catalytic C
[2a,17]
ꢀ
H activations and C C bond formations.
The desired
[3+2] product, 1-phenyl-1H-inden-1-amine (3a), was
formed by using catalysts generated from [{(cod)Rh(OH)}2]
(4; cod=1,5-cyclooctadiene) and phosphane ligands. An iso-
quinoline byproduct 5 was also detected,[11i] and reaction
conditions were screened to improve the yield of 3a. Under
the optimized conditions of 4 (0.5 mol%) and DPPP ligand
(1.2 mol%), in dry toluene at 1208C, the 1:1 reaction of 1a
is more commonly observed in tandem sequences initiated
[9]
ꢀ
by heteroatom-directed C H bond activation. For exam-
ple, Satoh, Miura, and co-workers have developed the Rh-
catalyzed oxidative coupling of benzoic acids with alkynes
and olefins to form isocoumarin and phthalide derivatives.[10]
Analogous syntheses of N-heterocycles have also been re-
ported by several groups.[11] In particular, a very recent
report by Satoh, Miura, and co-workers has shown that a
RhIII-based catalyst promoted indenone–imine formation by
oxidative coupling of N-arylbenzaldimines and alkynes in
Table 1. Development of the catalytic conditions.[a]
the presence of stoichiometric CuACTHNUTRGNENUG(OAc)2. However, the use
of ketimine substrates led to isoquinoline formation.[11i]
Therefore, a major challenge for our design is the selective
cyclization through intramolecular ketimine addition instead
of the formation of isoquinoline-type products (Scheme 1,
compound D).[5,11a,g–i] Recently, Kuninobu, Takai, and co-
workers have developed Re-based catalysts for a variety of
[3+2] annulations by a tandem sequence that is analogous
to the catalytic cycle shown in Scheme 1.[12] In particular,
Rh Catalyst
[{(cod)Rh(OH)}2] (4)
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Ligand[b]
Yield [%][c]
3a+5
3a/5
1
2
3
4
5
6
7
8
G
none
PEt3
PCy3
PtBu3
PPh3
PPh3
DPPM
DIPHOS
DPPP
DPPB
DPPpent
DPPF
XANTPHOS
BIPHEP
ACHTUNGTRENNUNG
<5
0
30
0
69
57
0
25
98
58
0
10
0
40
23
75
0
1:5
–
1:2.2
–
1:1.5
1:50[d]
–
5:1
50:1
10:1
–
2.5:1
–
1:1
2:1
7:1[e]
–
ꢀ
imine- or ketone-directed aromatic C H bond activation,
followed by alkyne or acrylate insertion, generated organo-
rhenium nucleophiles that underwent intramolecular ketone
or ketimine additions. Notably, except for one reported ex-
ample of N-benzyl tert-carbinamine,[12d] this rhenium cataly-
sis did not form isolatable tertiary carbinol or carbinamine
products due to the rapid loss of H2O/amine under the reac-
tion conditions. Thus, another challenge here is to retain the
newly formed tertiary carbinamine functionality, which pro-
vides the opportunity to develop relevant asymmetric cataly-
sis.[6]
9
10
11
12
13
14
15
16
17
18
19
4
ACHTUNGTRENNUNG
RhCl3·xH2O
A
DPPP
DPPP
DPPP
0
0
–
–
R
For the current study, we initially focused our attention
on relatively less studied, N-unsubstituted diaryl ketimine
substrates with the following considerations: First, diaryl ke-
timines can be prepared conveniently by Grignard reactions
by using readily available aryl halides and aryl cyanides.[13]
This modular synthetic approach would also allow us to ex-
plore reactivity dependence on various aromatic substituents
(see below). Second, [3+2] annulations with N-unsubstitut-
ed diaryl ketimines would be operationally simple and with-
out the need for additional deprotection procedures. Third,
reactions with diaryl ketimines will avoid potential interfer-
ence by imine/enamine tautomerization. It is noteworthy,
[a] Reaction conditions: 1a (0.23 mmol), 2a (1.05 equiv), Rh catalyst
(0.01 equiv Rh), phosphane (0.03 equiv for monophosphanes, 0.012 equiv
for chelating phosphanes), toluene (1.0 mL), 1208C, 18 h. [b] DPPM=
bis(diphenylphosphanyl)methane, DIPHOS=1,2-bis(diphenylphospha-
nyl)ethane, DPPP=1,3-bis(diphenylphosphanyl)propane, DPPB=1,4-
bis(diphenylphosphanyl)butane, DPPpent=1,5-bis(diphenylphosphanyl)-
pentane, DPPF=1,1’-bis-(diphenylphosphanyl)ferrocene, XANTPHOS=
4,5-bis(diphenylphosphanyl)-9,9-dimethylxanthene, BIPHEP=2,2’-bis(di-
phenylphosphanyl)-1,1’-biphenyl, BINAP=2,2’-bis(diphenylphosphanyl)-
1,1’-binaphthyl, (R,R)-DIOP=(4R,5R)-(ꢀ)4,5-bis(diphenylphosphanyl-
methyl)-2,2-dimethyl-1,3-dioxolane, COE=cyclooctene. [c] The com-
bined yields of 3a and 5 were determined by GC. [d] This reaction was
carried out in DMF and product 5 was isolated in 50% yield. [e] This re-
action was carried out with 2.5 mol% of 4. Product 3a was isolated in
65% yield and 51% ee (see the Supporting Information for HPLC analy-
sis).
ꢀ
however, that the relatively weak ketimine N H linkage
could lead to formation of metal–iminyl complexes. These
2620
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2619 – 2627

Tertiary Carbinamine Synthesis
FULL PAPER
Table 2. Scope of alkyne substrates for catalytic [3+2] annulation.[a]
and 2a gave a near-quantitative yield and over 50:1 selectiv-
ity of 3a versus 5 (Table 1, entry 9). The overall yield and
selectivity was significantly influenced by various factors,
such as rhodium precursor, phosphane ligand, and solvent
polarity. In particular, monophosphane ligands (Table 1, en-
tries 2–5) and polar solvents, such as DMF, CH3CN, and t-
amyl alcohol, led to lower yields and significantly lower se-
lectivities of 3a versus 5. The high efficiency of this catalyst
system encouraged us to attempt the enantioselective forma-
tion of 3a and up to 51% ee was achieved by using (R,R)-
DIOP as the chiral ligand (Table 1, entry 16). This prelimi-
nary result will serve as a proof-of-principle for the future
development of relevant processes in asymmetric catalysis.[6]
Notably, formation of 5 represents an isoquinoline synthe-
sis that draws parallels to recent studies by the groups of
Fagnou, Satoh, Miura, and others.[11] This formally dehydro-
genative [4+2] annulation could be improved by using PPh3
ligand and DMF solvent. However, several unidentified by-
products were also detected and 5 was isolated in 50% yield
despite a near-quantitative conversion and very high selec-
tivity of [4+2] versus [3+2] annulation (Table 1, entry 6).[18]
With the standard reaction conditions established, various
alkyne substrates were studied for Rh-catalyzed [3+2] an-
nulation with 1a (Table 2). Compared with diphenylacety-
lene (2a), ethyl- and n-propyl-substituted alkynes 2b, 2c,
and 2 f were slightly less reactive and required higher cata-
lyst loadings (1.0–1.5 mol% of 4) to give the annulation
products in high yields. In contrast, reactivities of methoxy-
methyl- and methyl-substituted alkynes 2d and 2e were sim-
ilar to that of 2a. Nonsymmetric alkynes 2e and 2 f reacted
with high regioselectivities to form the 2-phenyl-3-alkyl re-
gioisomers in >30:1 ratios. Among other alkyne substrates
Alkyne
Product
Rh loading Yield
[%]
[%][b]
1
2
3
1
94
2
3
88
83
4
5
6
1
1
2
90
88[c] (>50:1)
96[c] (>30:1)
[a] Reaction conditions: 1a (0.45 mmol),
2 (1.05 equiv), 4 (0.005–
0.015 equiv), DPPP (1.2 equiv per Rh), toluene (1.0 mL), 1208C, 18 h.
[b] Averaged yield of the isolated product from two runs. [c] Only the
major regioisomer was shown. The selectivity was determined by NMR
spectroscopic analysis.
ꢀ
that have been reported for catalytic C H alkenylations, ter-
bly, the reaction was severely inhibited by ortho substitu-
ents: no reactions occurred with 2-methoxybenzophenone
imine or 2,6-difluorobenzophenone imine, whereas ortho-
fluorinated 1j reacted reluctantly with higher catalyst load-
ing (55% yield with 2.5 mol% of 4). Interestingly, com-
pound 1j reacted with reversed regioselectivity that favored
coupling at the unsubstituted phenyl in a 2:1 ratio.
minal alkynes (e.g., phenylacetylene) and trialkylsilyl-substi-
tuted alkynes (e.g., bis(trimethylsilyl)acetylene and 1-
phenyl-2-trimethylsilylacetylene) were unreactive.[19] Car-
boxy alkynes, such as dimethyl- and diethyl acetylenedicar-
boxylates, decomposed under the current reaction condi-
tions.[20]
A series of diaryl ketimine substrates were studied for the
catalytic [3+2] annulation with 2a and the results are sum-
marized in Table 3. Diaryl ketimines with electron-with-
drawing F and CF3 groups at the para positions (1b, 1c, 1 f,
1g) were slightly less reactive than 1a, giving the [3+2]
products in high yields with 1.0 mol% loading of catalyst 4.
Similar reactivities were observed with substrates modified
by para-methoxy and meta-CF3 substituents (1h, 1i). How-
Current results on the substrate scope and effects of aro-
matic substituents provided the following mechanistic in-
sights into the proposed catalytic cycle (Scheme 1): First,
high regioselectivities for the incorporation of nonsymmetric
ꢀ
alkynes supported the proposed C H alkenylation via an ar-
ylrhodium(I) intermediate, A. As observed with products 3e
and 3 f, alkyne insertion into the RhI–aryl linkage placed the
Rh center preferentially at the benzylic position in the alke-
nylrhodium(I) intermediate, B (Scheme 1, A!B, R1 =alkyl,
R2 =Ar).[21] Second, alkyne insertion into a metal–hydro-
carbyl linkage is expected to be facilitated by p conjugation
and inhibited by steric hindrance. Therefore, the observed
higher overall reactivity with phenyl-substituted 2a and the
less sterically demanding alkynes 2d and 2e was consistent
with a rate-limiting alkyne insertion step. However, the re-
activity differences were not significant enough for a solid
conclusion. Third, moderate regioselectivities with nonsym-
metrical ketimines 1 f, 1g, and 1i argued against an electro-
ever, 3,3’-bisACHTUNGTRENNUNG(CF3)- and 4,4’-bisACHTUTGNREN(NUGN OMe)-substituted 1d and 1e
suffered low reactivity in toluene due to limited solubility,
giving <10% GC yields with standard reaction conditions.
Switching the solvent to DMF allowed 1d and 1e to give
good yields of [3+2] products with 1.5 mol% loading of 4,
although small amounts (<10%) of the isoquinoline by-
products were also detected. Nonsymmetrical, monosubsti-
tuted ketimines 1 f–i reacted with moderate regioselectivities
(ꢁ2–3:1) favoring coupling at the phenyl groups functional-
ized with para-F/CF3/methoxy and meta-CF3 groups. Nota-
Chem. Eur. J. 2010, 16, 2619 – 2627
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2621

P. Zhao et al.
Table 3. Scope of ketimine substrates for catalytic [3+2] annulation.[a]
Besides diaryl ketimines, sev-
eral types of structurally rele-
vant substrates were also ex-
plored for the catalytic [3+2]
annulation with alkynes. N-Un-
substituted monoaryl ketimines
appeared to be less reactive
than the diaryl analogues. For
example, valerophenone imine
Ketimine
Product
Yield [%][b]
94
1
Ar, Ar’=4-FC6H4 (1b)
(=NH))[9] was inert
(PhACTHNUGRTENNUG(nBu)CACHTUGNTRENNUGN
towards 2a with the current cat-
alyst system. In contrast, meta-
CF3-substituted valerophenone
imine 1l reacted smoothly with
2a to give the corresponding
[3+2] product 3q in 91% yield
(Scheme 3).[23] Reactions with
several aromatic ketones and
N-substituted aromatic imines
were also studied, but none of
them gave detectable amounts
of [3+2] products.[24]
2
3
4
Ar, Ar’=4-CF3C6H4 (1c)
Ar, Ar’=3-CF3C6H4 (1d)
Ar, Ar’=p-anisyl (1e)
83
74[c]
82[c]
The apparent lack of [3+2]
reactivity with aromatic ketones
and N-substituted imines was
quite intriguing, since these
substrates have been reported
to undergo several catalytic
5
6
Ar=4-FC6H4; Ar’=Ph (1 f)
Ar=4-CF3C6H4; Ar’=Ph (1g)
Ar=p-anisyl; Ar’=Ph (1h)
96 (2.4:1)
89 (3.2:1)
ꢀ
tandem C H bond activation/
alkyne insertion processes.[2,25–27]
This seemingly unique reactivi-
ty of N-unsubstituted ketimines
could result from facile forma-
tion of a ketimine–RhI s com-
plex,[28] or from formation of a
RhI–iminyl complex. Either
complex may serve as a reactive
intermediate to facilitate the ac-
7
8
95 (1.9:1)
92 (2.0:1)
Ar=3-CF3C6H4; Ar’=Ph (1i)
Ar=2-FC6H4; Ar’=Ph (1j)
9
55[d] (1:2)
ꢀ
tivation of aromatic C H
bonds. To explore such possibil-
ities, we have sought to prepare
RhI–imine and –iminyl com-
plexes and study their potential
roles in the proposed catalytic
cycle. Attempted reactions be-
[a] Reaction conditions:
1
(0.45 mmol), 2a (1.05 equiv),
4
(0.010 equiv), DPPP (1.2 equiv/Rh), toluene
(1.0 mL), 1208C, 18 h. [b] Averaged yield of isolated product from two runs. The selectivity of 3/3’ was deter-
mined by NMR spectroscopy and X-ray analysis (see the Supporting Information). [c] With 4 (0.015 equiv) in
DMF (1.0 mL) to improve solubility. [d] With 4 (0.025 equiv) and in toluene (1.5 mL).
ꢀ
philic substitution pathway for the aromatic C H activation
step, which would prefer alkenylation at electron-neutral
phenyl groups over ones with electron-withdrawing F and
CF3 substituents. As an additional mechanistic probe, we
synthesized 1k and studied its reaction with 2a (Scheme 2).
Consistent with a non-electrophilic substitution pathway,[22]
tween a DPPP-ligated rhodium(I) hydroxide and 1a failed
to form any detectable rhodium complexes (Scheme 4). In
ꢀ
the C H activation occurred exclusively at the meta-CF3-
substituted phenyl. Finally, the significantly reduced reactivi-
ty and reversed regioselectivity with ortho-fluorinated 1j
was consistent with involvement of the cyclometalated inter-
mediate A, which would be destabilized by ortho substitu-
ents due to steric hindrance.[2h]
Scheme 2. Reaction of imine 1k with 2a.
2622
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2619 – 2627

Tertiary Carbinamine Synthesis
FULL PAPER
to improvement of the catalyst efficiency for broader syn-
thetic applications and for asymmetric catalysis.
Scheme 3. Reaction of meta-CF3-substituted valerophenone imine 1l with
2a.
Experimental Section
General: Unless otherwise noted, all manipulations were carried out
under a nitrogen atmosphere by using standard Schlenk-line or glove box
techniques. All glassware was oven-dried for at least 1 h prior to use.
THF, diethyl ether, toluene, benzene, hexane, and pentane were degassed
by purging with nitrogen for 45 min and dried with a solvent purification
system (MBraun MB-SPS). CDCl3 and [D8]THF were degassed by purg-
ing with nitrogen and dried over activated 3 ꢁ molecular sieves. N-Un-
substituted diaryl ketimines were synthesized by Grignard reactions
based on a literature procedure.[13a] Other reagents and substrates were
purchased from commercial vendors and were used as received. TLC
plates were visualized by exposure to ultraviolet light or by exposure to
I2 sealed in a bottle at room temperature. Organic solutions were concen-
trated by rotary evaporation at ꢁ10 torr. Flash column chromatography
was performed with 32–63 micron silica gel.
Scheme 4. Attempted reaction between a DPPP-ligated rhodium(I) hy-
droxide and 1a.
GC analyses were carried out on Shimadzu GC-2010 with n-dodecane as
the internal standard. 1H NMR spectra were obtained on a 400 MHz
spectrometer and chemical shifts were recorded relative to residual proti-
ated solvent. 13C NMR spectra were obtained at 100 MHz and chemical
shifts were recorded relative to the solvent resonance. Both 1H and
13C NMR chemical shifts were reported in parts per million downfield
from tetramethylsilane (d=0 ppm). 31P NMR spectra were obtained at
121.5 MHz and chemical shifts were reported in parts per million down-
field of 85% H3PO4 (d=0 ppm). 19F NMR spectra were obtained at
282.4 MHz and all chemical shifts were reported in parts per million up-
field of CF3COOH (d=ꢀ78.5 ppm). HRMS were obtained on a Bruker
Daltronics BioTOF HRMS. HPLC analyses were carried out on a Waters
515 HPLC pump and a 2487 dual absorbance detector connected to a PC
with Empower workstation. Single-crystal X-ray diffraction data sets
were collected on a Bruker single-crystal X-ray diffractometer with a
SMART CCD 1 K area detector at 293 K with MoKa radiation. Crystals
were selected in paratone oil, mounted on a crystal loop, and positioned
in the diffractometer. The structure was solved and refined (on F2) by
using the SHELXL-97 package.[33] Please refer to the Supporting Infor-
mation for detailed results of single-crystal X-ray diffraction analysis on
compounds 3a, 3e, 3k/3k’, 5, and the rhodium(I) iminyl complex 6.
CCDC-743084, 743085, 743086, and 743088 contain the supplementary
crystallographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
contrast, metathesis between a DPPP-ligated rhodium(I)
chloride and the lithiated imine LiN=CPh2 did form a dis-
crete dimeric RhI–iminyl complex,[14] [{(dppp)Rh
ACTHNUTRGNE(NUG m-N=
CPh2)}2] (6), in 15% yield.[29] Complex 6 was fully character-
ized by NMR spectroscopy and single-crystal X-ray diffrac-
tion,[30] which supported a “bent-dimer” structure that was
commonly observed with dinuclear square planar complexes
of d8 metal centers.[31] Complex 6 showed high thermal sta-
bility upon heating at 1208C in toluene and no product from
ortho C H activation was detected by 31P NMR. Further-
ꢀ
more, compound 6 decomposed very slowly with added
alkyne 2a (5 equiv) upon heating at 1208C in toluene
(Scheme 5). Roughly 40% conversion was achieved after
Scheme 5. Stability of 6 at 1208C in the presence of 2a.
Preparation of 3-trifluoromethylvalerophenone imine (1l):[8] 3-Trifluoro-
methylbenzonitrile (621 mg, 3.63 mmol, 1.0 equiv) in THF (25 mL) was
added to a 100 mL round-bottomed flask equipped with a magnetic stir-
rer bar. The resulting solution was sealed and cooled to ꢀ788C; a balloon
filled with argon was fixed on the flask to balance the pressure. nBuLi
(3.3 mL, 1.6m in THF, 5.4 mmol, 1.4 equiv) was added dropwise at a rate
of 0.1 mLminꢀ1 (a syringe pump was applied), resulting in a red/brown
solution throughout the reaction. The solution was stirred at ꢀ788C for
one hour and an additional hour at room temperature and then treated
with small portions of Na2SO4·10H2O (0.59 g, 1.82 mmol, 0.5 equiv). The
mixture was stirred for 30 min during which time the color changed to
light yellow. The solvents were removed by rotary evaporation and the
residue was dissolved in 50 mL of EtOAc/Hexane (1:3) and then filtered
through Celite to remove inorganic salts. Crude yellow/brown oil
(680 mg, 82% yield) was obtained after removal of the solvents; further
purification by vacuum distillation gave 1l as a colorless oil. 1H NMR
(400 MHz, CDCl3): d=9.40 (very broad s, 1H), 8.01 (s, 1H), 7.92 (d, J=
6.0 Hz, 1H), 7.65 (d, J=7.6 Hz, 1H), 7.50 (t, J=8.0 Hz, 1H), 2.71 (t, J=
7.6 Hz, 2H), 1.58 (m, 2H), 1.35 (m, 2H), 0.95 ppm (t, J=7.6 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=177.7, 139.9, 130.0, 129.31 (q, J=
25.9 Hz), 129.18, 127.0, 124.1 (q, J=267 Hz), 123.7, 37.5, 28.1, 22.4,
14.0 ppm; 19F NMR (282.4 MHz, CDCl3): d=ꢀ63.3 ppm (s).
2 h, and only trace amounts of 3a and 5 were formed ac-
cording to NMR spectroscopy and GC analysis.[32] There-
fore, the preliminary mechanistic results are most consistent
with a catalytic cycle that does not directly involve the RhI–
iminyl intermediate 6, and a more systematic mechanistic in-
vestigation will be a focus for our future research.
Conclusion
A new method of Rh-catalyzed [3+2] annulation of N-un-
substituted aromatic ketimines and internal alkynes has
ꢀ
been developed based on imine-directed aromatic C H
bond activation and a rare example of intramolecular keti-
mine insertion into a RhI–alkenyl linkage. Current efforts
are focused on further mechanistic studies that would lead
Chem. Eur. J. 2010, 16, 2619 – 2627
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2623

P. Zhao et al.
General procedure for Rh-catalyzed [3+2] imine/alkyne annulations
(100 MHz, CDCl3): d=153.0, 150.4, 144.2, 142.8, 135.1, 134.6, 129.4,
128.5, 128.3, 127.8, 127.4, 126.8, 126.6, 125.8, 122.7, 119.6, 71.8, 11.9;
HRMS: m/z calcd for C22H17 (loss of -NH2 group): 281.1325; found:
281.1319.
Method A: The alkyne substrate (0.88 mmol, 1.0 equiv), imine substrate
(1.05 equiv), and a stock solution of Rh/DPPP in toluene (1.5 mL, con-
taining [{(cod)Rh(OH)}2] (0.005 equiv) and DPPP (0.012 equiv)) were
added to a 4 mL screw-cap vial equipped with a magnetic stirrer bar. The
vial was sealed with a silicone-lined screw cap, transferred out of the
glove box, and stirred at 1208C for 18 h. This reaction temperature and
reaction time was the same for Methods B–D. After the reaction mixture
was cooled to room temperature, all volatile materials were removed
under reduced pressure. Further purification was achieved by flash
column chromatography. Yields of the isolated products are based on the
average of two runs under identical conditions.
+
3-Ethyl-1,2-diphenyl-1H-inden-1-ylamine (3 f): Compound 3 f was pre-
pared from 2 f and 1a by using Method B. Chromatography (1:3 ethyl
acetate/hexane, Rf =0.40) gave 3 f as
a white solid (199 mg, 94%).
1H NMR (400 MHz, CDCl3): d=7.48–7.10 (m, 12H), 6.94–6.88 (m, 2H),
2.58 (m, 2H), 1.75 (s, -2H; NH2), 1.26 ppm (t, J=7.6 Hz, 3H); 13C NMR
(100 MHz, CDCl3): d=153.4, 150.3, 143.2, 142.5, 140.4, 135.2, 129.3,
128.4, 128.2, 127.7, 127.4, 126.8, 126.4, 125.9, 123.0, 120.0, 71.8, 19.5,
13.9 ppm; HRMS: m/z calcd for C23H19+: 295.1481; found: 295.1480.
5-Fluoro-1-(4-fluoro-phenyl)-2,3-diphenyl-1H-inden-1-ylamine
(3g):
Method B: Imine (0.68 mmol, 1.0 equiv), alkyne (1.05 equiv),
Compound 3g was prepared from 2a and 1b by using Method C. Chro-
matography (1:3 ethyl acetate/hexane, Rf =0.35) gave 3g as a white solid
(167 mg, 94%). This compound displayed the spectral patterns of unsym-
metrical phenyls due to limited bond rotations. 1H NMR (400 MHz,
CDCl3): d=7.50 (dd, J1 =8.4, J2 =5.2 Hz, 2H), 7.44–7.28 (m, 5H), 7.20–
7.02 (m, 4H), 6.98 (t, J=8.4 Hz, 3H), 6.85 (d, J=7.2 Hz, 3H), 1.81 ppm
[{(cod)Rh(OH)}2] (0.01 equiv), and DPPP (0.024 equiv).
Method C: Imine (0.45 mmol, 1.0 equiv), alkyne (1.05 equiv),
[{(cod)Rh(OH)}2] (0.015 equiv), and DPPP (0.036 equiv).
Method D: Reactions were performed in DMF (1.0 mL) with imine
(0.30 mmol,
1.0 equiv),
alkyne
(1.05 equiv),
[{(cod)Rh(OH)}2]
(0.015 equiv), and DPPP (0.036 equiv). The reaction mixture was cooled
and quenched with H2O (30 mL). The residue was extracted into CH2Cl2
(3ꢂ30 mL), washed with brine (3ꢂ20 mL), dried over anhydrous MgSO4,
filtered, and concentrated. Further purification was achieved by flash
column chromatography.
(s, 2H; -NH2); 13C NMR (100 MHz, CDCl3): d=163.2 (d,
242.7 Hz), 162.1(d, J(C,F)=244.1 Hz), 152.8, 148.2, 144.9 (d, J=8.1 Hz),
JACHTUNGTRENNUNG(C,F)=
A
ACHTUNGTRENNUNG
138.6, 138.3, 134.4, 134.0, 132.6 (d, J=8.1 Hz), 129.6, 129.4, 128.9, 128.2,
128.0, 127.7, 127.3 (d, J=8.1 Hz), 124.2 (d, J=9.5 Hz), 115.7, 115.5, 113.4,
113.1, 108.7, 108.5, 70.9 ppm; 19F NMR (282.4 MHz, CDCl3): d=ꢀ115.4
(t, J=9.3 Hz), ꢀ116.8 ppm (t, J=5.9 Hz); HRMS: m/z calcd for
1,2,3-Triphenyl-1H-inden-1-ylamine (3a): Compound 3a was prepared
from 2a and 1a by using Method A. Chromatography (1:4 ethyl acetate/
hexane, Rf =0.35) gave 3a as a white solid (210.0 mg, 87%). 1H NMR
(400 MHz, CDCl3): d=7.60–7.52 (m, 2H), 7.46–7.00 (m, 15H), 6.90–6.82
(m, 2H), 1.98 ppm (s, -NH2); 13C NMR (126 MHz, [D8]THF: d=154.3,
151.8, 143.8, 143.0, 139.2, 135.7, 135.2, 129.9, 129.6, 128.6, 128.3, 127.7,
127.5, 127.2, 126.9, 126.5, 126.4, 125.9, 123.2, 120.6, 72.1 ppm; HRMS: m/
+
C27H19NF2Na+: 418.1378; found: 418.1389; calcd for C27H17F2 (loss of
-NH2 group): 379.1293; found: 379.1300.
2,3-Diphenyl-5-trifluoromethyl-1-(4-trifluoromethylphenyl)-1H-inden-1-
ylamine (3h): Compound 3h was prepared from 2a and 1c by using
Method C. Chromatography (1:3 ethyl acetate/hexane, Rf =0.30) gave 3h
as a light yellow solid (185 mg, 83%). 1H NMR (400 MHz, CDCl3): d=
7.69 (dd, J1 =8.4, J2 =0.4 Hz, 2H), 7.56 (m, 3H), 7.50–7.32 (m, 6H), 7.28
(dd, J1 =8.0, J2 =0.4 Hz, 1H), 7.20–7.02 (m, 3H), 6.84 (dd, J1 =7.2, J2 =
0.4 Hz, 2H), 1.89 ppm (s, 2H; -NH2); 13C NMR (100 MHz, CDCl3): d=
155.8, 152.1, 146.4, 143.6, 139.2, 134.0, 133.5, 130.77 (q, J=32.4 Hz),
129.75 (q, J=32.0 Hz), 129.60, 129.4, 129.2, 128.43, 128.37, 128.0, 126.09,
125.99 (q, J=4.0 Hz), 124.39 (q, J=273.9 Hz), 124.29 (q, J=273.0 Hz),
124.1 (q, J=4.0 Hz), 123.5, 118.1 (q, J=4.0 Hz), 71.6 ppm; 19F NMR
(282.4 MHz, CDCl3): d=ꢀ62.65 (s), ꢀ62.97 ppm (s); HRMS: m/z calcd
for C29H17F6+: 479.1229; found: 479.1241.
+
z calcd for C27H19 (loss of -NH2 group): 343.1481; found: 343.1475.
2,3-Diethyl-1-phenyl-1H-inden-1-ylamine (3b): Compound 3b was pre-
pared from 2b and 1a by using Method B. Chromatography (1:2 ethyl
acetate/hexane, Rf =0.30) gave 3b as a colorless oil (157 mg, 88%).
1H NMR (400 MHz, CDCl3): d=7.35 (dt, J1 =6.8, J2 =0.8 Hz, 2H), 7.30–
7.16 (m, 5H), 7.12–7.05 (m, 2H), 2.66–2.50 (m, 2H), 2.34–2.10 (m, 2H),
1.76 (brs, 2H; -NH2), 1.27 (t, J=7.2 Hz, 3H), 0.89 ppm (t, J=7.6 Hz,
3H); 13C NMR (100 MHz, CDCl3): d=153.9, 151.6, 143.9, 142.9, 138.2,
128.2, 127.4, 126.7, 125.9, 125.5, 122.2, 118.8, 71.2, 18.8, 18.4, 14.7,
+
13.6 ppm; HRMS: m/z calcd for C19H19 (loss of -NH2 group): 247.1481;
2,3-Diphenyl-6-trifluoromethyl-1-(3-trifluoromethylphenyl)-1H-inden-1-
ylamine (3i): Compound 3i was prepared from 2a and 1d by using Meth-
od D. Chromatography (1:3 ethyl acetate/hexane, Rf =0.30) gave 3i as a
light yellow solid (165 mg, 74%). 1H NMR (400 MHz, CDCl3): d=7.94
(s, 1H), 7.63 (d, J=7.6 Hz, 1H), 7.53 (dd, J1 =8.0, J2 =0.8 Hz, 2H), 7.50–
7.30 (m, 8H), 7.20–7.00 (m, 3H), 6.82 (dd, J1 =7.2, J2 =1.2 Hz, 2H),
1.90 ppm (s, 2H; -NH2); 13C NMR (100 MHz, CDCl3): d=153.3, 152.8,
146.4, 143.3, 139.3, 134.2, 133.6, 131.4 (q, J=32.4 Hz), 129.6, 129.52,
129.47, 129.3, 129.09, 129.02 (q, J=32.0 Hz), 128.5, 128.3, 128.1, 125.8 (q,
J=4.1 Hz), 124.5 (q, J=4.1 Hz), 124.51 (q, J=272.0 Hz), 124.32 (q, J=
272.7 Hz), 122.6 (q, J=4.0 Hz), 121.5, 120.2 (q, J=4.1 Hz), 71.7 ppm;
19F NMR (282.4 MHz, CDCl3): d=ꢀ62.3 (s), ꢀ63.0 ppm (s); HRMS: m/z
calcd for C29H17F6+: 479.1229; found: 479.1228; calcd for C29H19NF6Na+:
518.1314; found: 518.1309.
found: 247.1480.
1-Phenyl-2,3-dipropyl-1H-inden-1-ylamine (3c): Compound 3c was pre-
pared from 2c and 1a by using Method C. Chromatography (1:2 ethyl
acetate/hexane, Rf =0.35) gave 2c as a colorless oil (109 mg, 83%).
1H NMR (400 MHz, CDCl3): d=7.31 (d, J=6.8 Hz, 2H), 7.28–7.12 (m,
5H), 7.05 (m, 2H), 2.49 (t, J=8.4 Hz, 2H), 2.30–2.14 (m, 1H), 2.14–1.98
(m, 1H), 1.76 (brs, 2H; -NH2), 1.74–1.60 (m, 2H), 1.42–1.22 (m, 1H),
1.20–1.05 (m, 1H), 1.02 (t, J=7.2 Hz, 3H), 0.82 ppm (t, J=7.2 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=153.9, 150.9, 144.1, 142.9, 137.0, 128.2,
127.4, 126.7, 125.9, 125.5, 122.2, 119.0, 71.2, 28.0, 27.8, 23.1, 22.1, 14.9,
+
14.6 ppm; HRMS: m/z calcd for C21H23 (loss of -NH2 group): 275.1794;
found: 275.1783.
2,3-Bis-methoxymethyl-1-phenyl-1H-inden-1-ylamine (3d): Compound
3d was prepared from 2d and 1a by using Method A. Chromatography
(3:2 ethyl acetate/hexane, Rf =0.35) gave 3d as a colorless oil (265 mg,
5-Methoxy-1-(4-methoxy-phenyl)-2,3-diphenyl-1H-inden-1-ylamine (3j):
Compound 3j was prepared from 2a and 1e by using Method D. Chro-
matography (1:2 ethyl acetate/hexane, Rf =0.40) gave 3j as a white solid
(102.6 mg, 82%). 1H NMR (400 MHz, CDCl3): d=7.60–7.25 (m, 7H),
7.20–7.00 (m, 4H), 6.95–6.80 (m, 5H), 6.71 (dd, J1 =8.0, J2 =0.8 Hz, 1H),
3.79 (s, 3H), 3.77 (s, 3H), 1.85 ppm (s, 2H; -NH2); 13C NMR (100 MHz,
CDCl3): d=160.0, 158.7, 152.6, 145.6, 144.4, 139.0, 135.20, 135.13, 134.7,
129.9, 129.6, 128.8, 128.2, 127.7, 127.4, 126.9, 123.9, 114.7, 111.7, 107.5,
70.9, 55.7, 55.4 ppm; HRMS: m/z calcd for C29H23O2: 403.1693; found:
403.1697.
1
90%). H NMR (400 MHz, CDCl3): d=7.45 (dt, J1 =7.6, J2 =0.8 Hz, 1H),
7.36–7.10 (m, 8H), 4.52 (s, 2H), 4.25 (d, J=12.4 Hz, 1H), 3.97 (d, J=
12.4 Hz, 1H), 3.42 (s, 3H), 3.23 (s, 3H), 2.03 ppm (s, 2H; -NH2);
13C NMR (100 MHz, CDCl3): d=153.0, 149.4, 142.3, 142.1, 136.7, 128.6,
128.0, 127.1, 127.0, 125.9, 122.8, 121.1, 71.5, 67.0, 66.3, 58.6, 58.5 ppm;
HRMS: m/z calcd for C19H21O2NNa+: 318.1465; found: 318.1455.
3-Methyl-1,2-diphenyl-1H-inden-1-ylamine (3e): Compound 3e was pre-
pared from 2e and 1a by using Method A. Chromatography (1:3 ethyl
acetate/hexane, Rf =0.40) gave 3e as a white solid (227 mg, 87%). The
assigned structure of 3e was confirmed by X-ray crystallography (please
5-Fluoro-1,2,3-triphenyl-1H-inden-1-ylamine (3k) and 1-(4-fluoro-
phenyl)-2,3-di-phenyl-1H-inden-1-ylamine (3k’): Compound 3k was pre-
pared from 2a and 1 f by using Method C. Chromatography (1:4 ethyl
acetate/hexane, Rf =0.30) gave a mixture of 3k and 3k’ as a white solid
1
see the Supporting Information for details). H NMR (400 MHz, CDCl3):
d=7.80–6.80 (m, 14H), 2.20 (s, 3H), 1.75 ppm (s, 2H; -NH2); 13C NMR
2624
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2619 – 2627

Tertiary Carbinamine Synthesis
FULL PAPER
(170 mg, 96%). Compounds 3k and 3k’ could not be separated by chro-
matography. Based on the analyses of single-crystal X-ray diffraction (see
the Supporting Information for details) and 19F NMR spectroscopy, 3k
55%, 3o:3o’=1:2). HRMS: m/z calcd for C27H18F+ (loss of -NH2 group):
361.1387; found: 361.1397. 3o: 1H NMR (400 MHz, CDCl3): d=7.47 (d,
1H), 7.42–7.16 (m, 12H), 7.10 (d, J=6.4 Hz, 1H), 7.08 (s, 1H), 7.04 (d,
J=6.4 Hz, 1H), 6.98–6.88 (m, 2H), 2.07 ppm (s, 2H; -NH2); 13C NMR
1
and 3k’ have a ratio of 71 to 29%. H NMR (400 MHz, CDCl3): d=7.80–
6.80 (m, 18H), 1.86 ppm (s, 2H; -NH2); 13C NMR (100 MHz, CDCl3): d=
163.3 (d, J=242.8 Hz), 162.2 (d, J=244.2 Hz), 153.2, 153.0, 151.0, 148.54,
148.51, 145.1 (d, J=8.1 Hz), 142.8, 142.7, 139.6, 138.73, 138.70, 135.1,
134.7, 134.5, 134.3, 132.7 (d, J=8.1 Hz), 129.8, 129.6, 129.5, 128.99,
128.97, 128.86, 128.3, 128.1, 127.9, 127.7, 127.58, 127.56 (d, J=8.1 Hz),
127.3, 127.0, 125.7, 124.4 (d, J=9.5 Hz), 123.3, 121.3, 115.8, 115.53, 113.4,
113.2, 108.7, 108.5, 71.5, 71.4 ppm; 19F NMR (282.4 MHz, CDCl3): d=
ꢀ114.8 (t, J=9.0 Hz), ꢀ116.3 ppm (t, J=9.0 Hz); HRMS: m/z calcd for
C27H18F+ (loss of -NH2 group): 361.1387; found: 361.1367.
(100 MHz, CDCl3): d=158.3 (d, JACTHNURTGNEU(GN F,C)=248.2 Hz), 151.7, 146.2, 141.1,
139.2, 138.1, 134.8, 134.1, 129.9, 129.7, 128.8, 128.7, 128.1, 128.0, 127.7,
127.4, 125.8, 122.6, 117.3, 114.4 (d, J=20.2 Hz), 71.4 ppm; 19F NMR
(282.4 MHz, CDCl3): d=ꢀ123.3 ppm (s). 3o’: 1H NMR (400 MHz,
CDCl3): d=7.95 (t, J=8.0 Hz, 1H), 7.60–6.80 (m, 17H), 1.94 ppm (s, 2H;
-NH2); 13C NMR (100 MHz, CDCl3): d=160.5 (d, J
ACTHNUTRGNE(NUG F,C)=248.2 Hz),
151.3, 149.2, 143.6, 139.6, 135.4, 134.8, 129.7, 129.6, 129.2, 129.1, 128.9 (d,
J=2.7 Hz), 128.7, 128.2, 125.1, 127.7, 127.4, 126.8, 124.4, 122.6, 121.2,
116.2 (d, J=20.5 Hz), 69.5 ppm; 19F NMR (282.4 MHz, CDCl3): d=
ꢀ113.7 (s).
1,2,3-Triphenyl-5-trifluoromethyl-1H-inden-1-ylamine (3l) and 2,3-di-
phenyl-1-(4-trifluoromethylphenyl)-1H-inden-1-ylamine (3l’): Compound
3l was prepared from 2a and 4-trifluoromethyl-benzophenone imine (1g)
by using Method C. Chromatography (1:4 ethyl acetate/hexane, Rf =0.40)
gave a mixture of 3l and 3l’ as a white solid (171 mg, 89%). Compounds
3l and 3l’ could not be separated by chromatography. Based on the anal-
ysis by 19F NMR spectroscopy, 3l and 3l’ have a ratio of 76 to 24%.
1H NMR (400 MHz, CDCl3): d=7.69 (d, 1H, J=8.0 Hz), 7.64–7.50 (m,
3H), 7.48–7.00 (m, 12H), 6.94–6.80 (m, 2H), 1.87 ppm (s, 2H; -NH2);
13C NMR (100 MHz, CDCl3): d=156.6, 152.8, 143.7, 141.8, 138.7, 134.4,
134.0, 130.40 (q, J=32.5 Hz), 129.8, 129.7, 129.6, 129.5, 129.12, 129.08,
128.9, 128.3, 128.2, 128.0, 127.9 (d, J=33.7 Hz), 127.7 (d, J=29.7 Hz),
127.1, 126.2, 125.9 (q, J=4.0 Hz), 125.7, 124.0 (q, J=4.6 Hz), 123.6, 123.3,
121.5, 117.9 (q, J=4.6 Hz), 71.8 ppm; 19F NMR (282.4 MHz, CDCl3): d=
1-(3-Methoxy-phenyl)-2,3-diphenyl-6-trifluoromethyl-1H-inden-1-ylamine
(3p): Compound 3p was prepared from 2a and 1k by using Method C.
Chromatography (1:4 ethyl acetate: hexane, Rf =0.30) gave 3p as a white
1
solid (173 mg, 84%). H NMR (400 MHz, CDCl3): d=7.52–7.45 (m, 2H),
7.42–7.30 (m, 6H), 7.30–7.16 (m, 2H), 7.14–7.06 (m, 4H), 6.89 (dt, J1 =
6.0, J2 =1.2 Hz, 2H), 6.82 (dq, J1 =8.0, J2 =0.8 Hz, 1H), 3.78 (s, 3H),
1.85 ppm (s, 2H; -NH2); 13C NMR (100 MHz, CDCl3): d=160.3, 153.7,
153.4, 146.4, 143.5, 138.7, 134.6, 133.9, 130.1, 129.8, 129.5, 129.0, 128.72
(q, J=31.2 Hz), 128.33, 128.16, 127.9, 125.4 (q, J=4.1 Hz), 124.6 (q, J=
272.7 Hz), 121.2, 120.1 (q, J=4.1 Hz), 118.0, 112.6, 111.8, 71.8, 55.4 ppm;
19F NMR (282.4 MHz, CDCl3): d=ꢀ62.2 ppm (s); HRMS: m/z calcd for
C29H20OF3+: 441.1461; found: 441.1478.
+
ꢀ62.7 (s), ꢀ63.0 ppm (s); HRMS: m/z calcd for C28H18F3 (loss of -NH2
1-Butyl-2,3-diphenyl-6-trifluoromethyl-1H-inden-1-ylamine (3q): Com-
pound 3q was prepared from 2a and 1l by using Method C. Chromatog-
raphy (1:4 ethyl acetate/hexane, Rf =0.40) gave 3q as a light yellow solid
group): 411.1355; found: 411.1381.
5-Methoxy-1,2,3-triphenyl-1H-inden-1-ylamine (3m) and 1-(4-methoxy-
phenyl)-2,3-diphenyl-1H-inden-1-ylamine (3m’): Compound 3m was pre-
pared from 2a and 1h by using Method C. Chromatography (1:1 ethyl
acetate/hexane, Rf =0.35) gave a mixture of 3m and 3m’ as a white solid
(166 mg, 95%). Compounds 3m and 3m’ could not be separated by chro-
matography. Based on the analyses by 1H NMR spectroscopy, 3m and
3m’ have a ratio of 65 to 35%. 1H NMR (400 MHz, CDCl3): d=7.62–
7.56 (m, 2H), 7.54–7.02 (m, 12H), 7.00–6.84 (m, 3H), 6.73 (dd, 1H, J1 =
8.4 Hz, J2 =2.4 Hz), 3.80, 3.79 (s, 3H), 1.93 ppm (s, 2H; -NH2); 13C NMR
(100 MHz, CDCl3): d=159.9, 158.6, 153.3, 152.4, 151.1, 145.3, 144.3,
143.2, 142.7, 139.1, 135.2, 135.0, 134.6, 134.5, 132.6, 131.9, 129.72, 129.68,
129.5, 128.71, 128.67, 128.0, 127.66, 127.60, 127.32, 127.25, 126.86, 126.81,
126.75, 125.6, 123.8, 123.1, 121.0, 114.1, 113.6, 111.6, 107.4, 71.3, 71.1,
55.6, 55.2 ppm; HRMS: m/z calcd for C28H21O+ (loss of -NH2 group):
373.1587; found: 373.1590.
1
(167 mg, 91%). H NMR (400 MHz, CDCl3): d=7.70 (q, J=0.8 Hz, 1H),
7.53 (dq, J1 =8.0, J2 =0.8 Hz, 1H), 7.35 (d, J=8.0 Hz, 1H), 7.32–7.20 (m,
10H), 2.08–1.80 (m, 2H), 1.64 (s, 2H), 1.24–1.04 (m, 4H), 0.76 ppm (t,
J=7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=152.3, 151.3, 146.9,
138.9, 135.2, 134.4, 129.62, 129.51, 128.66, 128.51, 128.35 (q, J=31.9 Hz),
127.85, 127.79, 124.9 (q, J=272.7 Hz), 125.2 (q, J=4.0 Hz), 120.8, 118.9
(q, J=4.0 Hz), 69.9, 38.2, 26.0, 22.9, 14.0 ppm; 19F NMR (282.4 MHz,
CDCl3): d=ꢀ62.1 ppm (s); HRMS: m/z calcd for C26H22F3+: 391.1668;
found: 391.1694; calcd for C26H24NF3Na+: 430.1753; found: 430.1778.
1,3,4-Triphenylisoquinoline (5): Compound 5 was prepared from 2a and
1a by using Method D with PPh3 as the ligand (benzophenone
(0.88 mmol), 1% Rh, PPh3/Rh (3 equiv)). Chromatography (1:4 ethyl
acetate/hexane, Rf =0.50) gave 5 as a light yellow solid (157 mg, 50%
yield). Crude 5 (60 mg) and Et2O (1 mL) were placed into a vial. The
mixture was sealed and heated at 808C for approximately 10 min until
completely dissolved. After cooling to room temperature, colorless crys-
tals were observed on the vial wall. Suitable crystals were selected under
a microscope for single-crystal X-ray diffraction (see the Supporting In-
formation for details). 1H and 13C NMR spectroscopy results were in
agreement with literature data.[11a]
1,2,3-Triphenyl-6-trifluoromethyl-1H-inden-1-ylamine (3n) and 2,3-di-
phenyl-1-(3-trifluoromethylphenyl)-1H-inden-1-ylamine (3n’): Compound
3n was prepared from 2a and 1i by using Method C. Chromatography
(1:4 ethyl acetate/hexane, Rf =0.40 for 3n, Rf =0.30 for 3n’) gave 3n
(65 mg), 3n’ (31 mg), and 3n+3n’ (80 mg) as white solids (total yield:
92%, 3n/3n’=2:1). 3n: 1H NMR (400 MHz, CDCl3): d=7.64–7.44 (m,
4H), 7.44–7.22 (m, 10H), 7.16–7.02 (m, 3H), 6.86 (m, 2H), 1.88 ppm (s,
2H; -NH2); 13C NMR (100 MHz, [D8]THF): d=155.1, 154.7, 146.9, 142.3,
138.2, 134.8, 134.5, 129.8, 129.4, 128.7, 128.5, 128.05 (quartet, J=32 Hz),
127.78, 127.70, 127.4, 127.0, 125.8, 125.0 (q, J=273 Hz), 124.7 (q, J=
4.6 Hz), 120.8, 119.8 (q, J=3.0 Hz), 72.2 ppm; 19F NMR (282.4 MHz,
CDCl3): d=ꢀ62.6 ppm (s). 3n’: 1H NMR (400 MHz, CDCl3): d=8.00 (s,
1H), 7.84–7.00 (m, 15H), 6.84 (s, 2H), 1.90 ppm (s, 2H; -NH2); 13C NMR
(100 MHz, [D8]THF): d=153.6, 151.0, 145.6, 142.8, 139.8, 135.3, 134.7,
130.45 (q, J=31.9 Hz), 129.8, 129.7, 129.4, 128.9, 128.6, 127.8, 127.59,
127.56, 127.1, 126.6, 123.4 (q, J=3.8 Hz), 123.1, 122.8 (q, J=4.6 Hz),
120.9, 71.9 ppm; 19F NMR (282.4 MHz, CDCl3): d=ꢀ63.4 ppm (s);
Synthesis and crystallization of [{(dppp)Rh
ACHTNUGRTEN(NNUG N=CPh2)}2] (6): [{(coe)2Rh-
AHCTUNGTRENNUNG
2.10 equiv), and toluene (3 mL) were added to a 4 mL vial (A), equipped
with a magnetic stirrer bar. The mixture was stirred for 20 min at RT to
form
a red solution of [{(dppp)RhACHTUNRTGENG(NUN m-Cl)}2]. LiNACHTUGNETRN(NUGN SiMe3)2 (30.9 mg,
0.185 mmol, 2.20 equiv), 1a (31.8 mg, 0.175 mmol, 2.10 equiv), and tolu-
ene (3 mL) were added to a 20 mL scintillation vial (B) equipped with a
magnetic stirrer bar. The mixture was stirred for 20 min at RT to form a
yellow solution of LiN=CPh2. [{(dppp)RhACTHNUTRGNEUNG(m-Cl)}2] in vial A was added
dropwise into vial B and then heated at 1008C for 20 min, at which time
>95% conversion was achieved based on 31P NMR spectroscopic analy-
sis. After filtration through Celite, the solution was concentrated to
2.5 mL under reduced pressure and layered by hexane (3.5 mL). Red/
brown block crystals were obtained on the vial wall after 3 d (25 mg,
18% yield based on Rh). 31P NMR (121.5 MHz, toluene): d=21.71 (d,
+
HRMS: m/z calcd for C28H18F3 (loss of -NH2 group): 411.1355; found:
411.1376.
7-Fluoro-1,2,3-triphenyl-1H-inden-1-ylamine (3o) and 1-(2-fluorophen-
yl)-2,3-diphenyl-1H-inden-1-ylamine (3o’): Compound 3o was prepared
from 2a and 1j by using Method C with 5% Rh catalyst. Chromatogra-
phy (1:4 ethyl acetate/hexane, Rf =0.40 for 3o, Rf =0.30 for 3o’) gave 3o
(22 mg), 3o’ (36 mg), and 3o+3o’ (38 mg) as white solids (total yield:
J
RhꢀP =159.7 Hz). This complex is extremely air sensitive and satisfactory
element analysis could not be obtained.
Chem. Eur. J. 2010, 16, 2619 – 2627
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2625

P. Zhao et al.
[8] For a recent example, see: T. Katagiri, T. Mukai, T. Satoh, K.
Hirano, M. Miura, Chem. Lett. 2009, 38, 118–119.
Acknowledgements
[9] For recent examples of heterocycle formation through Pd-catalyzed
domino sequences initiated by oxidative additions with aryl halides,
see reference [5] and D. A. Candito, M. Lautens, Angew. Chem.
2009, 121, 6841–6844; Angew. Chem. Int. Ed. 2009, 48, 6713–6716.
[10] a) K. Ueura, T. Satoh, M. Miura, Org. Lett. 2007, 9, 1407–1409;
b) T. Satoh, K. Ueura, M. Miura, Pure Appl. Chem. 2008, 80, 1127–
1134; c) M. Shimizu, K. Hirano, T. Satoh, M. Miura, J. Org. Chem.
2009, 74, 3478–3483.
Financial support for this work was provided by a ND EPSCoR seed
grant (EPS-0447679) and the NDSU start-up fund. We thank Dr. Angel
Ugrinov for assistance with the X-ray analysis and Anthony F. X. Pillai
for experimental assistance.
[1] L. F. Tietze, G. Brasche, K. Gericke, Domino Reactions in Organic
Synthesis, Wiley, New York, 2006.
[11] a) S.-G. Lim, J. H. Lee, C. W. Moon, J.-B. Hong, C.-H. Jun, Org.
Lett. 2003, 5, 2759–2761; b) D. A. Colby, R. G. Bergman, J. A.
Ellman, J. Am. Chem. Soc. 2006, 128, 5604–5605; c) D. A. Colby,
R. G. Bergman, J. A. Ellman, J. Am. Chem. Soc. 2008, 130, 3645–
3651; d) N. Umeda, H. Tsurugi, T. Satoh, M. Miura, Angew. Chem.
2008, 120, 4083–4086; Angew. Chem. Int. Ed. 2008, 47, 4019–4022;
e) L. Li, W. W. Brennessel, W. D. Jones, J. Am. Chem. Soc. 2008,
130, 12414–12419; f) D. R. Stuart, M. Bertrand-Laperle, K. M. N.
Burgess, K. Fagnou, J. Am. Chem. Soc. 2008, 130, 16474–16475;
g) N. Guimond, K. Fagnou, J. Am. Chem. Soc. 2009, 131, 12050–
12051; h) K. Parthasarathy, M. Jeganmohan, C.-H. Cheng, Org. Lett.
2008, 10, 325–328; i) T. Fukutani, N. Umeda, K. Hirano, T. Satoh,
M. Miura, Chem. Commun. 2009, 5141–5143.
[12] a) Y. Kuninobu, A. Kawata, K. Takai, J. Am. Chem. Soc. 2005, 127,
13498–13499; b) Y. Kuninobu, Y. Tokunaga, A. Kawata, K. Takai, J.
Am. Chem. Soc. 2006, 128, 202–209; c) Y. Kuninobu, Y. Nishina, M.
Shouho, K. Takai, Angew. Chem. 2006, 118, 2832–2834; Angew.
Chem. Int. Ed. 2006, 45, 2766–2768; d) Y. Kuninobu, Y. Nishina, K.
Okaguchi, M. Shouho, K. Takai, Bull. Chem. Soc. Jpn. 2008, 81,
1393–1401.
[2] For recent reviews, see: a) D. A. Colby, R. G. Bergman, J. A.
Ellman, Chem. Rev. 2009, DOI:10.1021/cr900005n; b) T. Kitamura,
Eur. J. Org. Chem. 2009, 1111–1125; c) F. Kakiuchi, T. Kochi, Syn-
thesis 2008, 3013–3039; d) B.-J. Li, S.-D. Yang, Z.-J. Shi, Synlett
2008, 949–957; e) D. Alberico, M. E. Scott, M. Lautens, Chem. Rev.
2007, 107, 174–238; f) T. Satoh, M. Miura, Top. Organomet. Chem.
2008, 24, 61–84; g) Y. J. Park, C.-H. Jun, Bull. Korean Chem. Soc.
2005, 26, 871–877; h) F. Kakiuchi, S. Murai, Acc. Chem. Res. 2002,
35, 826–834; i) V. Ritleng, C. Sirlin, M. Pfeffer, Chem. Rev. 2002,
102, 1731–1769; j) G. Dyker, Angew. Chem. 1999, 111, 1808–1822;
Angew. Chem. Int. Ed. 1999, 38, 1698–1712; k) Y. Guari, S. Sabo-
Etienne, B. Chaudret, Eur. J. Inorg. Chem. 1999, 1047–1055.
ꢀ
[3] a) This transformation from the catalyst (“LnM X”) to intermediate
A is shown in Scheme 1 as a single step, generating one equivalent
of HX. Such a one-step process is often used to describe an electro-
ꢀ
philic substitution pathway for C H bond activation. However, we
would like to clarify that the current one-step transformation is a
simplified description and does not imply an electrophilic substitu-
tion pathway (see the Results and Discussion for details). b) The
proposed intramolecular imine coordination in intermediate B is
likely to occur through a p complexation by the C=N moiety. As
one of the reviewers kindly pointed out, the alternative coordination
mode, s complexation by the nitrogen atom, would limit the rota-
tion of the electrophile (C=N moiety). This would disturb the orbital
interaction that is necessary for the proposed intramolecular imine
insertion and, as a result, inhibit the formation of the carbocycle.
For a related example with intramolecular coordination by the
imine nitrogen that led to the heterocycle formation, please see ref-
erence [11e].
[13] a) Y. Dejaegher, S. Mangelinckx, N. De Kimpe, Synlett 2002, 0113–
0115; b) G. Hou, F. Gosselin, W. Li, J. C. McWilliams, Y. Sun, M.
Weisel, P. D. OꢃShea, C.-Y. Chen, I. W. Davies, X. Zhang, J. Am.
Chem. Soc. 2009, 131, 9882–9883.
[14] For examples of b-carbon eliminations from RhI– and PdII–iminyl
species, see: a) T. Nishimura, S. Uemura, J. Am. Chem. Soc. 2000,
122, 12049–12050; b) P. Zhao, C. D. Incarvito, J. F. Hartwig, J. Am.
Chem. Soc. 2006, 128, 3124–3125; c) P. Zhao, J. F. Hartwig, Organo-
metallics 2008, 27, 4749–4757.
ꢀ
[15] For examples of C N bond formation through reductive elimina-
[4] For recent progress towards tert-carbinamines with catalytic or stoi-
chiometric use of organometallic reagents, see: a) G. K. Friestad,
A. K. Mathies, Tetrahedron 2007, 63, 2541–2569; b) M. Shibasaki,
M. Kanai, Chem. Rev. 2008, 108, 2853–2873; c) S. Kobayashi, H. Ko-
nishi, U. Schneider, Chem. Commun. 2008, 2313–2315; d) P. Fu,
M. L. Snapper, A. H. Hoveyda, J. Am. Chem. Soc. 2008, 130, 5530–
5541; e) A. D. Dilman, D. E. Arkhipov, V. V. Levin, P. A. Belyakov,
A. A. Korlyukov, M. I. Struchkova, V. A. Tartakovsky, J. Org. Chem.
2008, 73, 5643–5646; f) G. T. Notte, J. L. Leighton, J. Am. Chem.
Soc. 2008, 130, 6676–6677; g) A. M. Johns, Z. Liu, J. F. Hartwig,
Angew. Chem. 2007, 119, 7397–7399; Angew. Chem. Int. Ed. 2007,
46, 7259–7261.
tions from arylpalladium(II) iminyls, see: a) J. F. Hartwig, Acc.
Chem. Res. 1998, 31, 852–860; b) J. Barluenga, F. Aznar, C. Valdꢄs,
Angew. Chem. 2004, 116, 347–349; Angew. Chem. Int. Ed. 2004, 43,
343–345.
ꢀ
[16] For proposed C N bond formation through olefin insertions into a
Pd–iminyl linkage, see: a) S. Zaman, M. Kitamura, K. Narasaka,
Bull. Chem. Soc. Jpn. 2003, 76, 1055–1062; b) S. Zaman, M. Kita-
mura, A. D. Abell, Aust. J. Chem. 2007, 60, 624–626.
[17] a) Modern Rhodium-Catalyzed Organic Reactions (Ed.: P. A.
Evans), Wiley-VCH, Weinheim, 2005; b) K. Fagnou, M. Lautens,
Chem. Rev. 2003, 103, 169–196.
[18] It is likely that the low yields resulted from competitive pathways
mediated by rhodium hydride species. However, attempted reactions
with various added hydrogen acceptors did not give improved
yields.
[19] It is likely that terminal alkynes formed unproductive rhodium–al-
kynyl complexes, although detection of such species was not at-
tempted. The lack of reactivity of trialkylsilyl-substituted alkynes
could result from the steric hindrance or electronic effects from the
silyl functionality.
[5] For late-metal-catalyzed intramolecular ketone additions, see:
a) L. G. Quan, V. Gevorgyan, Y. Yamamoto, J. Am. Chem. Soc.
1999, 121, 3545–3546; b) L. G. Quan, M. Lamrani, Y. Yamamoto, J.
Am. Chem. Soc. 2000, 122, 4827–4828; c) L. Zhao, X. Lu, Angew.
Chem. 2002, 114, 4519–4521; Angew. Chem. Int. Ed. 2002, 41, 4343–
4345; d) G. Liu, X. Lu, J. Am. Chem. Soc. 2006, 128, 16504–16505;
e) T. Nishimura, Y. Yasuhara, T. Hayashi, J. Am. Chem. Soc. 2007,
129, 7506–7507; f) T. Miura, M. Murakami, Chem. Commun. 2007,
217–224.
[20] Styrene and n-butyl acrylate were also tested as olefin substrates,
[6] For an example of diastereoselective intramolecular N-sulfinyl keti-
mine insertion into a Pd–aryl linkage, see: Y.-B. Zhao, B. Mariampil-
lai, D. A. Candito, B. Laleu, M. Li, M. Lautens, Angew. Chem. 2009,
121, 1881–1884; Angew. Chem. Int. Ed. 2009, 48, 1849–1852.
[7] Mn-mediated stoichiometric intramolecular ketone and ketimine ad-
ditions have been observed, for examples, see: a) L. S. Liebeskind,
J. R. Gasdaska, J. S. McCallum, J. Org. Chem. 1989, 54, 669–677;
b) M. B. Dinger, L. Main, B. K. Nicholson, J. Organomet. Chem.
1998, 565, 125–134.
but neither showed any reactivity with the current catalyst system.
ꢀ
[21] An alternative and common C H alkenylation mechanism involves
generation of an arylrhodium
N
alkyne insertion into the RhIII–hydride linkage and subsequent C C
reduction elimination, see references [2a,h] for detailed discussions.
ꢀ
ꢀ
[22] Plausible pathways for the directed C H bond activation include ox-
ꢀ
idative addition and s-bond metathesis. In particular, C H oxidative
III
ꢀ
addition of “LnM X” would generate a Rh intermediate; subse-
2626
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 2619 – 2627

Tertiary Carbinamine Synthesis
FULL PAPER
quent loss of HX through reductive elimination would generate the
proposed arylrhodium(I) intermediate A as shown in Scheme 1 (see
ref. [2] for detailed discussions). For a recent mechanism study, see:
L. Li, W. W. Brennessel, W. D. Jones, Organometallics 2009, 28,
3492–3500.
reactivity with Lewis acidic RhI or RhIII precursors argued against
this possibility (Table 1, entries 17–19). For a toluene sulfonic acid
(TsOH)-catalyzed cyclization of ortho-alkenyl aromatic ketones,
see: P. W. R. Harris, P. D. Woodgate, Synth. Commun. 1997, 27,
4195.
[23] The CF3-induced rate enhancement could result from the higher re-
[28] a) M. B. Ezhova, B. O. Patrick, B. R. James, Organometallics 2005,
24, 3753–3757; b) M. L. Buil, M. A. Esteruelas, E. Goni, M. Olivan,
E. Onate, Organometallics 2006, 25, 3076–3083; c) for an example
of an aromatic ketone–rhodium(I) complex, see: C. P. Lenges, M.
Brookhart, J. Am. Chem. Soc. 1999, 121, 6616–6623.
[29] This preparation proceeded with high chemical yields (>80%); the
low yield of the isolated product reflected the solubility issue during
crystallization.
ꢀ
activity of C H activation with an electron-poor aryl, or from the
higher reactivity of intramolecular addition to an electron-poor
imine electrophile.
[24] Attempted substrates include acetophenone, benzophenone, N-
phenyl benzophenone imine, N-tert-butyl benzaldimine, and benzo-
phenone O-pentafluorobenzoyl oxime.
[25] a) U. R. Aulwurm, J. U. Melchinger, H. Kisch, Organometallics 1995,
14, 3385–3395; b) C.-H. Jun, J.-B. Hong, Y.-H. Kim, K.-Y. Chung,
Angew. Chem. 2000, 112, 3582–3584; Angew. Chem. Int. Ed. 2000,
39, 3440–3442; c) H. Werner, N. Mahr, J. Wolf, A. Fries, M. Lau-
bender, E. Bleuel, R. Garde, P. Lahuerta, Organometallics 2003, 22,
3566–3576; d) S.-G. Lim, J.-A. Ahn, C.-H. Jun, Org. Lett. 2004, 6,
4687–4690; e) K. Tsuchikama, Y. Kuwata, Y.-K. Tahara, Y. Yoshina-
mi, T. Shibata, Org. Lett. 2007, 9, 3097–3099.
[30] See the Supporting Information for details.
[31] a) M. A. Esteruelas, F. J. Lahoz, M. Olivan, E. Onate, L. A. Oro, Or-
ganometallics 1994, 13, 3315–3323; b) G. Aullꢅn, S. Alvarez, Inorg.
Chem. 2001, 40, 4937–4946; c) W. J. Marshall, G. Aullon, S. Alvarez,
K. D. Dobbs, V. V. Grushin, Eur. J. Inorg. Chem. 2006, 3340–3345.
[32] GC analysis was carried out after the reaction was quenched with
dry MeOH. The major product was the recovered benzophenone
imine 1a.
[26] a) For Rh-catalyzed ortho-arylation of N-unsubstituted benzophe-
ꢀ
none imine with NaBPh4 through imine-directed C H bond activa-
[33] G. M. Sheldrick, Crystallographic Software Package, SHELXTL,
version 5.1, Bruker-AXS, Madison, WI, 1998.
tion, see: K. Ueura, T. Satoh, M. Miura, Org. Lett. 2005, 7, 2229–
2231; b) see reference [8] for a comparison of the reactivity of N-un-
substituted ketimines and N-silyl analogues.
[27] Lewis acidic metal complexes may also promote the cyclization of
ortho-alkenyl aromatic imine intermediates. However, the lack of
Received: October 12, 2009
Published online: January 14, 2010
Chem. Eur. J. 2010, 16, 2619 – 2627
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2627