Catalytic formal [2+2+1] synthesis of pyrroles from alkynes and diazenes via TiII/TiIV redox catalysis

Article abstract of DOI:10.1038/nchem.2386

Pyrroles are structurally important heterocycles. However, the synthesis of polysubstituted pyrroles is often challenging. Here, we report a multicomponent, Ti-catalysed formal [2+2+1] reaction of alkynes and diazenes for the oxidative synthesis of penta- and trisubstituted pyrroles: a nitrenoid analogue to classical Pauson-Khand-type syntheses of cyclopentenones. Given the scarcity of early transition-metal redox catalysis, preliminary mechanistic studies are presented. Initial stoichiometric and kinetic studies indicate that the mechanism of this reaction proceeds through a formally Ti^II/Ti^IV redox catalytic cycle, in which an azatitanacyclobutene intermediate, resulting from [2+2] alkyne + Ti imido coupling, undergoes a second alkyne insertion followed by reductive elimination to yield pyrrole and a Ti^II species. The key component for catalytic turnover is the reoxidation of the Ti^II species to a Ti^IV imido via the disproportionation of an η²-diazene-Ti^II complex.

Full text of DOI:10.1038/nchem.2386

ARTICLES
PUBLISHED ONLINE: 2 NOVEMBER 2015 | DOI: 10.1038/NCHEM.2386
Catalytic formal [2+2+1] synthesis of pyrroles
from alkynes and diazenes via TiII/TiIV
redox catalysis
Zachary W. Gilbert, Ryan J. Hue and Ian A. Tonks
*
Pyrroles are structurally important heterocycles. However, the synthesis of polysubstituted pyrroles is often challenging.
Here, we report a multicomponent, Ti-catalysed formal [2+2+1] reaction of alkynes and diazenes for the oxidative
synthesis of penta- and trisubstituted pyrroles: a nitrenoid analogue to classical Pauson–Khand-type syntheses of
cyclopentenones. Given the scarcity of early transition-metal redox catalysis, preliminary mechanistic studies are
presented. Initial stoichiometric and kinetic studies indicate that the mechanism of this reaction proceeds through a
formally TiII/TiIV redox catalytic cycle, in which an azatitanacyclobutene intermediate, resulting from [2+2] alkyne + Ti
imido coupling, undergoes a second alkyne insertion followed by reductive elimination to yield pyrrole and a TiII species.
The key component for catalytic turnover is the reoxidation of the TiII species to a TiIV imido via the disproportionation of
an η2-diazene-TiII complex.
he development of earth-abundant and nontoxic catalysts for the sub-stoichiometric oxidative formation of pyrroles during
precision chemical transformations is an imperative endea- Ti-catalysed alkyne hydroamination reactions, which were generated
T
vour of modern synthesis, and recent years have seen a great via reductive elimination of expanded metallacycles to yield TiII
effort towards replacing the precious group 9 and 10 metal catalysts species32,33. Based on this precedent, we were interested in closing
with their lighter base metal congeners1. Complementary to this a potential catalytic cycle by reoxidation of TiII to TiIV = NR with
research, many synthetically practical processes that use earth-abun- a nitrene oxidant34. Given several reports of low-valent Ti species
dant early transition-metal or lanthanide catalysts have also been cleaving diazenes to imidos35–37, we were led to explore the reactivity
developed2–12. However, due to the oxophilic, electropositive of azobenzene, which may serve as a weakly nucleophilic nitrene
nature of early transition metals, the majority of organometallic cat- source and simple Ti imidos, (L)nTiCl2(NR) (L = HNMe2,py;
t
alytic processes that use them are redox-neutral4–6. Examples of cat- n = 2, 3; R = Ph, Tol, Bu)38,39, with various alkynes.
alytic processes that rely on early transition-metal redox processes
Here, we report the (py)3TiCl2(NR)-catalysed synthesis of pyr-
are rare and mostly limited to C–C7–11 or C–H12 bond-forming reac- roles through a three-component formal [2+2+1] oxidative coupling
tions, the most notable class being Kulinkovich-type cyclopropana- of alkynes and diazenes (Fig. 1b). This methodology represents a
tions8. Instead of proceeding through oxidative processes, early new retrosynthetic disconnection in the catalytic synthesis of
transition-metal-catalysed C–N bond-forming reactions typically pyrroles, and also demonstrates a unique example of a catalytic
occur through redox-neutral alkene/alkyne hydroamination5 and oxidative nitrene transfer by a TiII/TiIV redox couple.
related pathways6,7, and the only example of catalytic oxidative
C–N bond formation is a report from Heyduk that uses a Results and discussion
Zr complex with a redox non-innocent ligand to catalytically Reaction of 10 mol% 1a with 2 equiv. 3-hexyne (2a) and 0.5 equiv.
couple isonitriles and azides to form carbodiimides13.
azobenzene (3a) in mesitylene at 180 °C for 24 h yielded the cataly-
Polysubstituted pyrroles play a key role in pharmaceuticals14, tic production of N-phenyl-2,3,4,5-tetraethylpyrrole (4a) (22%), the
materials15, dyes16 and natural products17, and are often challenging formal [2+2+1] oxidative coupling product of two alkynes and 0.5
synthetic targets18–21. Various heterocycles have previously been azobenzenes (Fig. 2, equation (1)). Additionally, stoichiometric
formed from early transition metallacycles, but these reactions amounts (10%) of the hydroamination product, N-phenylhexan-
are predominantly limited to stoichiometric reactivity22–24. A cat- 3-imine, and small amounts of the cyclotrimerization product, hex-
alytic [2+2+1] cyclization involving nitrenes25 and alkynes would aethylbenzene, were observed. Although the yield and selectivity of
be closely related to the Pauson–Khand synthesis of cyclopente- this initial reaction are poor, three observations can be made. First,
nones, and adapting such well-established synthetic methods26 to azobenzene is capable of turning over the catalytic reaction through
nitrenes would allow for the rapid development of complex func- reoxidation of low-valent Ti intermediates. Second, the [2+2+1]
tionalized pyrroles27–29
.
reaction is competing with hydroamination, and the two reaction
Two reports have previously demonstrated the stoichiometric cycles probably share a common Ti intermediate. Third, TiII inter-
oxidative formation of pyrroles from alkynes by Ti complexes mediates must be involved due to the presence of competitive alkyne
(Fig. 1). First, Rothwell30 reported that bis(aryloxide)titanacyclo- trimerization10. In this case, hydroamination occurs via double pro-
pentadienes can insert benzo[c]cinnoline into a Ti–C bond to form tonation of an azatitanacyclobutene intermediate by the 2 equiv.
a diazatitanacycloheptadiene, which on heating in the presence of dimethylamine present in the Ti imido precatalyst.
excess benzo[c]cinnoline undergoes ring contraction to generate a
By using an aprotic catalyst, (py)3TiCl2(NPh) (1b)39, quantitative
Ti imido and a tethered pyrrole. More recently, Bercaw31 reported conversion to the desired [2+2+1] product was observed with
Department of Chemistry, University of Minnesota – Twin Cities, 207 Pleasant St SE, Minneapolis, Minnesota 55455, USA. e-mail: itonks@umn.edu
*
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
1
© 2015 Macmillan Publishers Limited. All rights reserved

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2386
ARTICLES
a
Rothwell
b
This work
R'
Et
R
Et
[2+2+1]
N
R
R
Et
ArO
2
+
“NR'”
OAr
Et
N
Et
N
N
Et
Et
ArO
ArO
N
Ti
Ti
N
R
N
R
R
N
N
Ti
N
ArO
ArO
Et
Ring
contraction
Et
Et
Et
N
OAr = 2,6-diphenylphenoxide
0.5 R'N=NR'
R'
N
+
Et
R
R
R
Cat. [Ti]
Formal [2+2+1]
Bercaw
2
Me
R
R
Me
Me
R
Ph
N
R = alkyl, aryl, TMS, H
R' = aryl
Ph
Me
Me
Ph
Me
Me
[(ONO)TiII]
Me
N
Me
N
Me
+
Alkyne
trimerization
catalyst
(ONO)Ti
(ONO)Ti
Reductive
elimination
Me
Me
ONO = pyridine-2,6-bis(4,6-tBu2-phenolate)
Figure 1 | Ti-mediated oxidative pyrrole formation. a, Previous stoichiometric examples of oxidative pyrrole formation from alkynes by Ti complexes.
b, Retrosynthetic disconnection and forward reaction for catalytic formal [2+2+1] oxidative cyclization of alkynes and diazenes.
minimal competing cyclotrimerization and no hydroamination production of 0.5 equiv. azobenzene, consistent with the η2-azoben-
(Fig. 2, equation (2)). Interestingly, both NPh units in azobenzene zene adduct being a viable catalytic intermediate.
are incorporated into the product. When using 10 mol% (py)3TiCl2-
Complex 7 could disproportionate to 1a through two limiting path-
(Ntol) (1c) or (py3)TiCl2(NtBu) (1d), 10% of the product contains ways: (1) by dimerization48 to generate a Ti2N4 species that could then
tolyl- or tBu-functionalized pyrrole, indicating that reactivity occurs retrocyclize into 1a and free azobenzene (Fig. 3b, top pathway) or (2)
through the imido functionality and that virtually all of the Ti catalyst by decoordination of azobenzene to make a free TiII species that could
undergoes reaction.
then comproportionate36 with another equivalent of 7 to generate
The preliminary mechanism for the catalytic oxidative formal 2 equiv. 1a (Fig. 3b, bottom pathway). The disproportionation is
[2+2+1] cyclization of alkynes and diazenes is presented in likely to occur through the dimerization pathway, because addition
Fig. 3a. First, a TiIV imido (I) undergoes [2+2] addition with an of tolNNtol to 7 yields neither tolyl-functionalized Ti imido nor
alkyne to form an azatitanacyclobutene (II), identical to the tolNNPh crossover products upon heating and disproportionation.
first step for Ti-catalysed hydroamination5,40–42. This step is
Notably, later transition metals and complexes with redox-active
supported by the initial catalytic experiments with 1a, where ligands have been reported to perform the reverse reaction—coup-
hydroamination competed with [2+2+1] cyclization. Next, a ling metal imidos to generate diazenes49–52. For this [2+2+1] coup-
second equivalent of alkyne inserts into the azatitanacyclobutene ling, the large thermodynamic preference for TiII oxidation drives
to form an azatitanacyclohexadiene (III). Second insertions the reaction to favour diazene cleavage to form imidos. The use of
into azatitanacycles are quite rare, but have previously been diazenes to promote Ti reoxidation is critical for catalysis. Other
observed by Mountford32 (alkynes, stoichiometric) and Odom ‘nitrene’ oxidants such as aryl azides or PhINTs do not yield
(isocyanides, catalytic)43.
productive reactivity.
After second insertion, reductive elimination from III yields the
Having determined the optimum reaction conditions (Supple-
pyrrole product and a TiII intermediate (IV). The resulting TiII mentary Sections 1 and 2), the scope of the reaction was examined
intermediate can then either unproductively trimerize alkyne10, or using a number of alkynes at 110 °C in TFT. The results of reactions
be trapped by azobenzene to form a TiII η2-azobenzene44–47
,
with internal and terminal alkynes, diynes and enynes are shown in
adduct V, which then disproportionates35–37 into a TiIV imido (I) Table 1. Overall, internal and terminal alkynes with alkyl, aryl and
to close the catalytic cycle. This reductive elimination may be facili- silyl groups are well tolerated, but alkynyl esters, tethered alkyl
tated by the coordination of azobenzene or alkyne, either of which ethers and bis(trimethylsilyl)acetylene do not react under
could act as a ‘redox-non-innocent’ ligand and immediately accept optimized conditions (see Supplementary Section 3).
the electrons from reductive elimination to avoid a long-lived TiII
Alkynes that are known to rapidly undergo cyclotrimerization
intermediate. Under either scenario, the binding competition of (such as terminal alkynes) require excess azobenzene to give high
azobenzene and alkyne determines whether alkyne trimerization yields. For example, phenylacetylene (2g) gave almost no pyrrole
or productive reoxidation occurs.
products (<5%) under standard conditions; instead, the major pro-
The reoxidation of the transient TiII species to a TiIV imido is ducts observed were 1,2,4-triphenylbenzene and 1,3,5-triphenylben-
proposed to involve a disproportionation of an η2-azobenzene zene. Increased yields of the pyrrole regioisomers 5g and 6g can be
complex into a TiIV imido and 0.5 equiv. azobenzene. To examine obtained by using a sixfold excess of 3a to favour azobenzene
the stoichiometric viability of this rather unique catalytic reoxida- binding over phenylacetylene binding to the TiII intermediate IV.
tion mechanism,
a
Ti η2-azobenzene complex, (HNMe2)2
Further evidence for the binding competition of alkyne and azo-
TiCl2(PhNNPh) (7), which is similar in structure to the proposed benzene can be observed in the reactions of tethered diynes.
TiII η2-azobenzene adduct V, was synthesized via protonolysis of Reaction of 2,7-nonadiyne (2k) with stoichiometric azobenzene
TiCl2(NMe2)2 by 1,2-diphenylhydrazine (Fig. 4). The X-ray resulted in 22% 4k, with the remainder of the 2k being converted
structure of 7 reveals an N–N bond length of 1.420(3) Å, indicat- to the tethered diaryl trimer that results from three diynes cyclotri-
ing a reduced η2-PhN-NPh(2−) hydrazido unit bound to a TiIV merizing. Increasing the concentration of azobenzene significantly
centre instead of a TiII azobenzene adduct37,47. Thermolysis of 7 improved the yield of 4k. Conversely, reaction of 3,9-dodecadiyne
in α,α,α-trifluorotoluene (TFT) results in rapid, full conversion to (2m) with only one equivalent of azobenzene gave a high yield for
the TiIV imido 1a (confirmed by 1H NMR), with concomitant the fused pyrrole 4m, highlighting that subtle changes in ring size
2
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2015 Macmillan Publishers Limited. All rights reserved

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2386
ARTICLES
Ph
N
Et
Conversely, hydroamination of phenylacetylene (2g) is less selec-
tive, yielding a 1:2 mixture of the imine products resulting from
protonolysis of A and B, respectively (Supplementary Section 8b).
During [2+2+1] catalysis with 2g, products arising from metalla-
cycles D, E and F are observed. In this case it is likely that the
initial [2+2] step is similarly unselective as in hydroamination,
and the observed product distribution is a result of moderately selec-
tive insertion of 2g into A and B, where the 2,1-addition leading to
metallacycle C is disfavoured over the other insertion pathways.
Interestingly, 2d yields an unselective mixture of all three poss-
ible pyrrole regioisomers, yet hydroamination is completely selective
for the methyl benzyl imine product that results from protonolysis
of only one of the metallacycles. In this case, the [2+2] reaction must
be reversible on the timescale of second alkyne insertion, allowing
for unselective secondary insertion of the sterically and electronically
poorly differentiated sides of the alkyne, yielding a mixture of
all products.
Et
Et
Et
Et
Et
Et
Et
2
+
Et
10%
2a
Et
Et
4a
(HNMe2)2Cl2Ti(NPh) (1a)
Et
(1)
+
Formal [2+2+1]
(22%)
Trimerization
(trace)
Mesitylene, 180 °C, 24 h
NPh
H
0.5 PhN=NPh
+
Et
H
3a
Et
Hydroamination
(10%)
Et
2
Ph
N
Et
Et
10%
Et
Et
Et
2a
Et
py3Cl2Ti(NPh) (1b)
(2)
+
+
°C, 16 h
PhCF3, 110
Et
Et
Et
Et
0.5 PhN=NPh
4a
Et
3a
Formal [2+2+1]
(quantitative)
Trimerization
(trace)
a
0.5 RN=NR
Diazene
disproportionation
R'
N
LnTiIV
N
Figure 2 | Initial catalytic experiments reveal that aprotic Ti imido
complexes are highly efficient catalysts for [2+2+1] cyclization.
LnTiIV
N
R
V
RN=NR
R'
R1
I
can shift the competition to favour product formation, even at low
azobenzene concentrations.
Alkyne
trimerization
R2
Because the mechanisms of first and second alkyne insertions are
different ([2+2] addition versus migratory insertion), there is great
potential to expand catalysis to include the selective coupling of two
different unsaturated substrates. For example, compared to alkynes,
[LnTiII]
[2+2] addition
IV
R
R
LnTiIV
R2
alkenes will not facilely undergo [2+2] reaction with Ti imidos5,41
,
N
R1
N
R4
R
R1
but will rapidly undergo migratory insertion; thus it may be possible
to catalytically couple alkynes and alkenes with Ti imidos. As
expected, tethered enyne 2n undergoes alkyne [2+2] addition fol-
lowed by alkene insertion to generate an azatitanacyclohexene.
However, instead of reductive C–N bond formation to yield a dihydro-
pyrrole, the metallacycle rearranges before reductive elimination to
yield the α,β-unsaturated imine 4n.
Reductive
elimination
R1
R3
II
N
LnTiIV
R2
R3
R2
Insertion
R4
R3
III
R4
Independent diazene adduct synthesis and thermal disproportionation
b
Coupling of untethered unsymmetrical alkynes can potentially
yield mixtures of regioisomers. Regioselectivity is determined
during two steps of catalysis (Fig. 4): (1) by the orientation of the
alkyne during [2+2] addition to give one of two regioisomeric aza-
metallacyclobutenes (A and B) and (2) during alkyne insertion,
which can proceed via 1,2- or 2,1-insertion into either A or B, ulti-
mately yielding one of four regioisomeric azametallacyclohexa-
dienes (C, D, E or F) that then undergo reductive elimination to
give the three possible pyrrole regioisomers 4, 5 and 6. Regiocontrol
is highly substrate-dependent (Table 1): 2d and 2f are essentially
non-selective for all three possible regioisomers; phenylacetylene
(2g) is moderately selective for the 2,4-disubstituted regioisomer
5g (4g:5g:6g = 0:4.0:1.0); and the terminal and trimethylsilyl
(TMS)-protected alkynes 2e, 2h, 2i and 2j are highly selective for
regioisomer 5.
Ph
N
Cl
Cl
TiCl2(NMe2)2
+
Rapid
C6H6
Me2HN Ti NHMe2
Me2HN Ti NHMe2
H
Ph
Δ
N
N
Cl
Cl
N
N
Ph
Ph
1a
+ 0.5 PhN=NPh
7
Ph
H
Potential disproportionation mechanisms
[Ti]
[Ti]
“retro-[2+2+2]”
– PhNNPh
Dimerization
PhN
NPh
NPh
Ph
N
7
PhN
Cl
Cl
Me2HN Ti NHMe2
2
Me2HN Ti
NHMe2
N
N
Cl
Cl
Ph
Ph
1a
7
7
The origin of selectivity in the [2+2+1] couplings of unsym-
metric alkynes can be qualitatively determined by comparing the
[2+2+1] regioselectivity to the selectivity of alkyne hydroamination
“[TiII]”
– PhNNPh
Azobenzene
decoordination
Comproportionation
catalysed by 1b, as both reactions share common [2+2] regio- Figure 3 | Mechanism of Ti-catalysed formal [2+2+1] oxidative cyclization
isomeric metallacycle intermediates (A and B)53. Hydroamination of alkynes and diazenes. a, The reaction is proposed to proceed through a
t
of BuCCH (2i) is completely selective for the sterically preferred TiII/TiIV redox couple. A TiIV imido (I) undergoes [2+2] addition with an alkyne
anti-Markovnikov imine product resulting from protonolysis of to form an azatitanacyclobutene (II). Next, a second equivalent of alkyne
metallacycle B (Supplementary Section 8c). During [2+2+1] catalysis inserts into the azatitanacyclobutene to form an azatitanacyclohexadiene (III).
with 2i, only 5i, which could arise from either metallacycles D or E, Reductive elimination from III yields the pyrrole product and a TiII intermediate
is observed. Based on the selectivity for B observed in the initial (IV). The resulting TiII intermediate can then either unproductively trimerize
[2+2] step, it is likely that 5i results only from sterically preferred alkyne, or be trapped by azobenzene to form a TiII η2-azobenzene adduct (V),
2,1-insertion of alkyne into B to yield metallacycle D. In the cases which then disproportionates into a TiIV imido (I) to close the catalytic cycle.
where 5 is the predominant regioisomeric product, selectivity for b, Synthesis and potential mechanisms for the disproportionation of 7 to 1a.
both [2+2] and second insertion are controlled by alkyne sterics.
Experimental evidence indicates that the dimerization pathway is more likely.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
3
© 2015 Macmillan Publishers Limited. All rights reserved

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2386
ARTICLES
Insertion
Ph
R2
Ph
N
N
N
N
R1
R2
R2
R2
R1
R2
Unselective (4, 5 and 6) substrates
A and B mixture/formation reversible
Second insertion unselective
R1
R1
R2
R2
LnTi
R2
kAC
2,1
1,2
R1
R1
R2
R1
R2
R2
R1
R2
4
Ph
R2
C
D
Me
Me
Ph
iPr
LnTi
R1
N
[2+2] addition
Ph
R1
R2
kAD
A
Semi-selective (5 and 6) substrates
Mixture of metallacycles A and B
kA
LnTi
R1
Ph
N
k–A
kAC disfavoured
R2
R1
R1
H
Ph
TiLn
N
Ph
Ph
Selective (5 only) substrates
Steric control
Metallacycle B dominates
kBE ≫ kBF
k–B
5
kB
LnTi
Ph
R1
kBE
R1
R2
LnTi
R2
N
2,1
1,2
R2
R1
R1
H
tBu
E
F
[2+2] addition
H
nBu
Ph
B
Ph
N
kBF
N
TMS
H
R1
LnTi
nPr
TMS
R2
R1
R1
6
R2
Insertion
Figure 4 | Mechanistic scheme explaining selectivity differences between various unsymmetric alkynes. The orientation of the alkyne during [2+2]
addition results in two possible regioisomeric azametallacyclobutenes: A and B. The alkyne insertion, which can proceed via 1,2- or 2,1-insertion into either A
or B, ultimately yields one of four regioisomeric azametallacyclohexadienes (C, D, E or F) that then undergo reductive elimination to give the three possible
pyrrole regioisomers 4, 5 and 6.
Table 1 | Initial alkyne scope of Ti-catalysed formal oxidative [2+2+1] cyclization.
Ph
N
Ph
N
Ph
N
R1
R1
R2
R2
R2
R1
10% py3Cl2Ti(NPh) (1b)
3 R1
R
2 + 0.5 PhN=NPh
+
+
PhCF3, 110 oC, 16 h
R2
6d–6j
R2
R1
4a–4n
R1
5d–5j
R1
R2
2a–2r
3a
when R1 = R2
Product
Entry Alkyne
Product(s) (ratio)
%Yield (NMR)* Entry Alkyne
%Yield (NMR)*
Et
Et
1†
4a
85 (96)
2a
Ph
N
11
2k
4k
4l
(76)
Me
Me
Ph
2
4b
76 (98)
2b
Ph
N
Ph
Ph
Ph
Ph
3
12
46
26 (51)
2l
4c
2c
Ph
Et
iPr
TMS
Ph
Me
4
4d (0.72):5d (1.0):6d (0.35)
4e (0.0):5e (1.0):6e (0.0)
4f (0.45):5f (1.0):6f (0.77)
4g (0.0):5g (1.0):6g (0.25)
4h (0.13):5h (1.0):6h (0.13)
4i (0.0):i (1.0):6i (0.0)
30 (71)
2d
Et
nPr
Ph
13
2m
2n
N
4m 65 (95)
5
(78)
2e
Et
Et
Me
6
60 (91)
2f
nBu
nBu
50|| (83)
14
4n
Ph
7‡
(32)
2g
N
nBu
15
2o
2p
No reaction
No reaction
8‡
O
O
36 (57)
2h
16
tBu
9
55 (92)
2i
TMS
MeO2C
TMS
17
18
2q
2r
No reaction
No reaction
TMS
10‡
4j (min)§:5j (max)§:6j (min)§
(52)
2j
CO2Me
Conditions: 1.21 mmol 2 (6 equiv.), 0.20 mmol 3 (1 equiv.), 0.020 mmol 1b (0.1 equiv.), 1 ml CF3Ph, 16 h; *NMR yield based on 3a with Ph3CH internal standard; †apparent turnover frequency based on initial
rates is ∼1 h−1 ‡6 equiv. 3a; yield based on 2; §overlapping 1H NMR signals prevent accurate integration: >80% 5h; ||isolated yield upon hydrolysis to the corresponding ketone.
;
4
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2015 Macmillan Publishers Limited. All rights reserved

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2386
ARTICLES
have revealed that the regioselectivity of the alkyne coupling is under
substrate control. However, given recent successes in ligand develop-
ment for selective and functional group-tolerant early transition-
metal-catalysed Pauson–Khand54 and hydroamination5,55 reactions,
it may be possible to apply similar principles to advance substrate
scope and selectivity in the catalytic formal [2+2+1] reaction.
Table2 | Initial diazene scope of Ti-catalysed formal
oxidative [2+2+1] cyclization.
R
N
10% py3Cl2Ti(NPh) (1b)
Et
Et
3
Et
Et + 0.5 RN=NR'
PhCF3, 140 °C, 16 h
Et
Et
Received 10 June 2015; accepted 29 September 2015;
published online 2 November 2015
2a
3b–3i
4
Entry Diazene
Product
%Yield (NMR)*
References
N
1. Bullock, R. M. Catalysis without Precious Metals (Wiley, 2010).
2. Mikami, K., Terada, M. & Matsuzawa, H. ‘Asymmetric’ catalysis by lanthanide
complexes. Angew. Chem. Int. Ed. 41, 3554–3571 (2002).
3. Nishiura, M. & Hou, Z. Novel polymerization catalysts and hydride clusters from
rare-earth metal dialkyls. Nature Chem. 2, 257–268 (2010).
4. Dudnik, A. S., Weidner, V. L., Motta, A., Delferro, M. & Marks, T. J.
Atom-efficient regioselective 1,2-dearomatization of functionalized
pyridines by an earth-abundant organolanthanide catalyst. Nature Chem. 6,
1100–1107 (2014).
5. Müller, T. E., Hultzsch, K. C., Yus, M., Foubelo, F. & Tada, M.
Hydroamination: direct addition of amines to alkenes and alkynes. Chem. Rev.
108, 3795–3892 (2008).
6. Odom, A. New C–N and C–C bond forming reactions catalyzed by titanium
complexes. Dalton Trans. 225–233 (2005).
7. Basuli, F., Wicker, B., Huffman, J. C. & Mindiola, D. J. Understanding the role of
an easy-to-prepare aldimine-alkyne carboamination catalyst,
[Ti(NMe2)3(NHMe2)][B(C6F5)4]. J. Orgomet. Chem. 696, 235–243 (2011).
8. Kulinkovich, O. G. The chemistry of cyclopropanols. Chem. Rev. 103,
2597–2632 (2003).
9. Hicks, F. A., Kablaoui, N. M. & Buchwald, S. L. Scope of the intramolecular
titanocene-catalyzed Pauson–Khand type reaction. J. Am. Chem. Soc. 121,
5881–5898 (1999).
10. Ozerov, O. V., Patrick, B. O. & Lapido, F. T. Highly regioselective [2+2+2]
cycloaddition of alkynes catalyzed by η6-arene complexes of titanium supported
by dimethylsilyl-bridged p-tert-butyl calix[4]arene ligand. J. Am. Chem. Soc.
122, 6423–6431 (2000).
11. Negishi, E., Swanson, D. R., Cederbaum, F. E. & Takahashi, T. Zirconium-
promoted bicyclization of enynes. Effects of enyne structure. Tetrahedron Lett.
28, 917–920 (1987).
3b
4ab R = 4-MePh
85 (86)
1
N
N
N
2
3c
3d
4ac R = 2-MePh
71 (83)
(32)
N
N
N
tBu
tBu
4ad R = 3,5-di-tBuPh
3
tBu
tBu
MeO
N
N
4
3e
3f
4ae R = 4-MeOPh
66 (86)
OMe
CF3
F3C
5
No reaction
N
N
3g
3h
4ac R = Bn
Trace†
6
N
N
O
12. Al Dulayymi, J. R., Baird, M. S., Bolesov, I. G., Tveresovsky, V. & Rubin, M.
A simple and efficient hydrodehalogenation of 1,1-dihalocyclopropanes.
Tetrahedron Lett. 37, 8933–8936 (1996).
13. Nguyen, A. I., Zarkesh, R. A., Lacy, D. C., Thorson, M. K. & Heyduk, A. F.
Catalytic nitrene transfer by a zirconium(IV) redox-active ligand complex. Chem.
Sci. 2, 166–169 (2011).
14. Baumann, M., Baxendale, I. R., Ley, S. V. & Nikbin, N. An overview of the key
routes to the best selling 5-membered ring heterocyclic pharmaceuticals.
Beilstein J. Org. Chem. 7, 442–495 (2011).
i
PrO
N
OiPr
7
No reaction
O
N
8
3i
4a:4ai (1:1)
(69)
N
NMe2
15. Vernitskaya, T. V. & Efimov, O. N. Polypyrrole: a conducting polymer;
its synthesis, properties, and applications. Russ. Chem. Rev. 66,
443–457 (1997).
Conditions: 1.21 mmol 2 (6 equiv.), 0.20 mmol 3 (1 equiv.), 0.020 mmol 1b (0.1 equiv.), 1 ml CF3Ph,
16 h; *NMR yield based on 3a with Ph3CH internal standard; †Bn2N2 undergoes radical
decomposition under the reaction conditions.
16. Loudet, A. & Burgess, K. BODIPY dyes and their derivatives: syntheses and
spectroscopic studies. Chem. Rev. 107, 4891–4932 (2007).
17. Walsh, C. T., Tsodikova, S. G. & Howard-Jones, A. R. Biological formation of
pyrroles: nature’s logic and enzymatic machinery. Nat. Prod. Rep. 23,
517–531 (2006).
18. Estévez, V., Villacampa, M. & Menéndez, J. C. Recent advances in the
synthesis of pyrroles by multicomponent reactions. Chem. Soc. Rev. 43,
4633–4657 (2014).
19. Martin, R., Larsen, C. H., Cuenca, A. & Buchwald, S. L. Cu-catalyzed tandem
C–N bond formation for the synthesis of pyrroles and heteroarylpyrroles. Org.
Lett. 9, 3379–3382 (2007).
20. Ramanathan, B., Keith, A. J., Armstrong, D. & Odom, A. L. Pyrrole syntheses
based on titanium-catalyzed hydroamination of diynes. Org. Lett. 6,
2957–2960 (2004).
Further substrate scope analysis was performed using substituted
aryl diazenes (Table 2). Tolyl-substituted diazenes 3b and 3c reacted
similarly to the parent azobenzene; however, more sterically encum-
bered diazene 3d was lower yielding. Electron-rich diazene 3e, which
contains a heteroatom-functionalized arene, also gives excellent con-
version to 4ae. Of synthetic importance, the N-methoxyphenyl sub-
stituent in 4ae can potentially be oxidatively deprotected to yield
the N-unsubstituted pyrrole. Finally, using an asymmetric diazene
(3i) yields both products 4a and 4ai in a 1:1 ratio, indicating both
halves of the diazene are incorporated into the pyrroles at a
roughly equal rate.
In summary, a simple Ti imido precatalyst, py3Cl2Ti(NPh), is
capable of performing inter- and intramolecular oxidative coupling
of alkynes or enynes with diazenes to generate polysubstituted pyr-
roles. This reaction is a unique example of catalytic oxidative C–N
bond formation with a group 4 early transition metal and of cataly-
tic formal [2+2+1] six-electron cyclization with a nitrene feedstock.
A key feature of the catalytic cycle is the cleavage of the N=N double
bond of an aryl diazene, which occurs through the disproportiona-
tion of a TiII η2-diazene adduct to reform a TiIV imido. Initial studies
21. Gulevich, A. V., Dudnik, A. S., Cernyak, N. & Gevorgyan, V. Transition metal-
mediated synthesis of monocyclic aromatic heterocycles. Chem. Rev. 113,
3084–3213 (2013).
22. Yu, S., Xiong, M., Xie, X. & Lui, Y. Insertion of nitriles into zirconocene
1-aza-1,3-diene complexex: chemoselective synthesis of N–H and N-substituted
pyrroles. Angew. Chem. Int. Ed. 53, 11596–11599 (2014).
23. Nakamoto, M. & Tilley, T. D. Reactions of zirconacyclopentadienes with
nitrosobenzene. Characterization of zirconacycle intermediates and formation of
N-phenylpyrroles. Organometallics 20, 5515–5517 (2001).
24. Rosenthal, U., Burlakov, V. V., Bach, M. A. & Beweries, T. Five-membered
metallacycles of titanium and zirconium—attractive compounds for
organometallics chemistry and catalysis. Chem. Soc. Rev. 36, 719–728 (2007).
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
5
© 2015 Macmillan Publishers Limited. All rights reserved

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2386
ARTICLES
25. Liu, W., Jiang, H. & Huang, L. One-pot silver-catalzyed and PIDA-mediated
sequential reactions: synthesis of polysubstituted pyrroles directly from
alkynoates and amines. Org. Lett. 12, 312–315 (2010).
26. Blanco-Urgoiti, J., Añorbe, L., Pérez-Serrano, L. & Domínguez, G. The
Pauson–Khand reaction, a powerful synthetic tool for the synthesis of
complex molecules. Chem. Soc. Rev. 33, 32–42 (2004).
27. Omae, I. Three characteristic reactions of alkynes with metal compounds in
organic synthesis. Appl. Organometal. Chem. 22, 149–166 (2008).
28. Chopade, P. R. & Louie, J. [2+2+2] cycloaddition reactions catalyzed by
transition metals. Adv. Synth. Catal. 348, 2307–2327 (2006).
29. Wender, P. A. et al. Inspirations from nature. New reactions, therapeutic leads,
and drug delivery systems. Pure Appl. Chem. 75, 143–155 (2003).
30. Hill, J. E., Fanwick, P. E. & Rothwell, I. P. Formation of a terminal aryl-imido
compound of titanium by cleavage of the N=N double bond in benzo[c]
cinnoline. Inorg. Chem. 30, 1143–1144 (1991).
31. Tonks, I. A., Meier, J. C. & Bercaw, J. E. Titanium complexes supported by
pyridine-bis(phenolate) ligands: active catalysts for intermolecular
hydroamination or trimerization of alkynes. Organometallics 32,
3451–3457 (2013).
32. Vujokvic, N. et al. Imido-alkyne coupling in titanium complexes: new
insight into the alkyne hydroamination reaction. Organometallics 26,
5522–5534 (2007).
33. Lokare, K. S., Ciszewski, J. T. & Odom, A. L. Group-6 imido activation by a ring-
strained alkyne. Organometallics 23, 5386–5388 (2004).
34. Fout, A. R., Kilgore, U. J. & Mindiola, D. J. The recent progression of synthetic
strategies to assemble titanium complexes bearing the terminal imide group.
Chem. Eur. J. 13, 9428–9440 (2007).
35. Duchateau, R., Williams, A. J., Gambarotta, S. & Chiang, M. Y. Carbon-carbon
double-bond formation in the intermolecular acetonitrile reductive coupling
promoted by a mononuclear titanium(II) compound. Preparation and
characterization of two titanium(IV) imido derivatives. Inorg. Chem. 30,
4863–4866 (1991).
44. Tripepi, G., Young, V. G. Jr & Ellis, J. E. Highly reduced organometallics Part 49.
Reaction of hexacarbonyltitanate(2–) with azobenzene. Structural
characterization of the first hydroxo-carbonyl of titanium [Ti2(μ-OH)2(CO)8]2−
J. Organomet. Chem. 593–594, 354–360 (2000).
.
45. Kaleta, K., Arndt, P., Spannenberg, A. & Rosenthal, U. Unusual bond activation
processes in the reaction of group 4 cyclopentadienyl alkyne complexes with
azobenzene. Inorg. Chim. Acta 370, 187–190 (2011).
46. Kaleta, K. et al. Reactions of group 4 metallocene alkyne complexes with
azobenzene: formation of diazametallacyclopropanes and N=N bond activation.
Organometallics 29, 2604–2609 (2010).
47. Retbøll, M. & Jørgensen, K. A. MO explanation of the structures of azo-
transition metal complexes. Inorg. Chem. 33, 6403–6405 (1994).
48. Goetze, B., Knizek, J., Noth, H. & Schnick, W. 1,2-Bis(trimethylsilyl)hydrazido
titanium complexes. Eur. J. Inorg. Chem. 1849–1854 (2000).
49. Munhá, R. F., Zarkesh, R. A. & Heyduk, A. F. Group transfer reactions of d0
transition metal complexes: redox-active ligands provide a mechanism for
expanded reactivity. Dalton Trans. 42, 3751–3766 (2013).
50. Zarkesh, R. A., Ziller, J. W. & Heyduk, A. F. Four-electron oxidative formation of
aryl diazenes using a tantalum redox-active ligand complex. Angew. Chem. Int.
Ed. 47, 4715–4718 (2008).
51. Mankad, N. P., Müller, P. & Peters, J. C. Catalytic N–N coupling of aryl azides to
yield azoarenes via trigonal bipyramid iron–nitrene intermediates. J. Am. Chem.
Soc. 132, 4083–4085 (2010).
52. Mansuy, D., Battioni, P. & Mahy, J. P. Isolation of an iron–nitrene complex from
dioxygen- and iron poryhrin-dependent oxidation of a hydrazine. J. Am. Chem.
Soc. 104, 4487–4489 (1982).
53. Gräbe, K., Pohlki, F. & Doye, S. Neutral Ti complexes as catalysts for the
hydroamination of alkynes and alkenes: do the labile ligands change the catalytic
activity? Eur. J. Org. Chem. 28, 4815–4823 (2008).
54. Hicks, F. & Buchwald, S. L. Highly-enantioselective catalytic Pauson–Khand
type formation of bicyclic cyclopentenones. J. Am. Chem. Soc. 118,
11688–11689 (1996).
55. Yim, J. C. H., Bexrud, J. A., Ayinla, R. O., Leitch, D. C. & Schafer, L. L.
Bis(amidate)bis(amido) titanium complex: a regioselective intermolecular
alkyne hydroamination catalyst. J. Org. Chem. 79, 2015–2028 (2014).
36. Gray, S. D., Thorman, J. L., Adamian, V. A., Kadish, K. M. & Woo, L. K.
Synthesis, electrochemistry, and imido and transfer reactions of (TTP)Ti(η2-
PhN=NPh). Inorg. Chem. 37, 1–4 (1998).
37. Hill, B. J. E., Profile, R. D., Fanwick, P. E. & Rothwell, Z. P. Synthesis, structure,
and reactivity of aryloxo(imido)titanium complexes. Angew. Chem. Int. Ed. Engl.
102, 664–665 (1990).
38. Adams, N. et al. New titanium imido synthons: syntheses and supramolecular
structures. Inorg. Chem. 44, 2882–2894 (2005).
39. Blake, A. J. et al. Synthesis and imido-group exchange reactions of
tert-butylimidotitanium complexes. J. Chem. Soc. Dalton Trans.
1549–1558 (1997).
Acknowledgements
Financial support was provided by the University of Minnesota (start-up funds).
Equipment purchases for the Chemistry Department NMR facility were supported by a
grant from the National Institutes of Health (S10OD011952) with matching funds from the
University of Minnesota. The Bruker-AXS D8 Venture diffractometer was purchased
through a grant from NSF/MRI (1224900) and the University of Minnesota.
Author contributions
40. Vujkovic, N. et al. Insertions into azatitanacyclobutenes: new insights into three-
component coupling raections involving imidotitanium intermediates.
Organometallics 27, 2518–2528 (2008).
41. Straub, B. F. & Bergman, R. G. The mechanism of hydroamination of
allenes, alkynes, and alkenes catalyzed by cyclopentadienyltitanium–imido
complexes: a density functional study. Angew. Chem. Int. Ed. 40,
4632–4635 (2001).
42. Weitershaus, K. et al. Titanium hydroamination catalysts bearing a
2-aminopyrrolinto spectator ligand: monitoring the individual reaction steps.
Dalton Trans. 4586–4602 (2009).
Z.W.G. and I.A.T. conceived and designed the experiments. Z.W.G. and R.J.H. performed
the experiments and analysed the data. I.A.T. wrote the manuscript. All authors
contributed to revising the manuscript.
Additional information
Supplementary information and chemical compound information are available in the
online version of the paper. Reprints and permissions information is available online at
www.nature.com/reprints. Correspondence and requests for materials should be
addressed to I.A.T.
43. Barnea, E., Majumder, S., Staples, R. J. & Odom, A. L. One step route to
2,3-diaminopyrroles using a titanium-catalyzed four-component coupling.
Organometallics 28, 3876–3881 (2009).
Competing financial interests
The authors declare no competing financial interests.
6
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2015 Macmillan Publishers Limited. All rights reserved