Mechanism of Ti-Catalyzed Oxidative Nitrene Transfer in [2 + 2 + 1] Pyrrole Synthesis from Alkynes and Azobenzene

Article abstract of DOI:10.1021/jacs.8b03546

A combined computational and experimental study on the mechanism of Ti-catalyzed formal [2 + 2 + 1] pyrrole synthesis from alkynes and aryl diazenes is reported. This reaction proceeds through a formally Ti^II/Ti^IV redox catalytic cycle as determined by natural bond orbital (NBO) and intrinsic bond orbital (IBO) analysis. Kinetic analysis of the reaction of internal alkynes with azobenzene reveals a complex equilibrium involving Ti=NPh monomer/dimer equilibrium and Ti=NPh + alkyne [2 + 2] cycloaddition equilibrium along with azobenzene and pyridine inhibition equilibria prior to rate-determining second alkyne insertion. Computations support this kinetic analysis, provide insights into the structure of the active species in catalysis and the roles of solvent, and provide a new mechanism for regeneration of the Ti imido catalyst via disproportionation. Reductive elimination from a 6-membered azatitanacyclohexadiene species to generate pyrrole-bound Ti^II is surprisingly facile and occurs through a unique electrocyclic reductive elimination pathway similar to a Nazarov cyclization. The resulting Ti^II species are stabilized through backbonding into the π? of the pyrrole framework, although solvent effects also significantly stabilize free Ti^II species that are required for pyrrole loss and catalytic turnover. Further computational and kinetic analysis reveals that in complex reactions with unysmmetric alkynes the resulting pyrrole regioselectivity is driven primarily by steric effects for terminal alkynes and inductive effects for internal alkynes.

Full text of DOI:10.1021/jacs.8b03546

Subscriber access provided by UNIV OF DURHAM
Article
On the Mechanism of Ti-Catalyzed Oxidative Nitrene Transfer
in [2+2+1] Pyrrole Synthesis from Alkynes and Azobenzene
Zachary W Davis-Gilbert, Xuelan Wen, Jason D. Goodpaster, and Ian A. Tonks
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b03546 • Publication Date (Web): 15 May 2018
Downloaded from http://pubs.acs.org on May 15, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 20
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
On the Mechanism of Ti-Catalyzed Oxidative Nitrene Transfer in
[2+2+1] Pyrrole Synthesis from Alkynes and Azobenzene
Zachary W. Davis-Gilbert,‡ Xuelan Wen,‡ Jason D. Goodpaster* and Ian A. Tonks*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Department of Chemistry, University of Minnesota – Twin Cities, Minneapolis, MN 55455
ABSTRACT: A combined computational and experimental study on the mechanism of Ti-catalyzed formal [2+2+1] pyrrole
synthesis from alkynes and aryl diazenes is reported. This reaction proceeds through a formally TiII/TiIV redox catalytic
cycle as determined by natural bond orbital (NBO) and intrinsic bond orbital (IBO) analysis. Kinetic analysis of the
reaction of internal alkynes with azobenzene reveals a complex equilibria involving Ti=NPh monomer/dimer equilibrium,
Ti=NPh + alkyne [2+2] cycloaddition equilibrium, along with azobenzene and pyridine inhibition equilibria prior to rate
determining 2nd alkyne insertion. Computations support this kinetic analysis and provide insights in the structure of the
active species in catalysis, the roles of solvent, and provide a new mechanism for regeneration of the Ti imido catalyst via
disproportionation. Reductive elimination from a 6-membered azatitanacyclohexadiene species to generate pyrrole-
bound TiII is surprisingly facile, and occurs through a unique electrocyclic reductive elimination pathway similar to a
Nazarov cyclization. The resulting TiII species are stabilized through backbonding into the π* of the pyrrole framework,
although solvent effects also significantly stabilize free TiII species that are required for pyrrole loss and catalytic turnover.
Further computational and kinetic analysis reveals that in complex reactions with unysmmetric alkynes, the resulting
pyrrole regioselectivity is driven primarily by steric effects for terminal alkynes and inductive effects for internal alkynes.
1. INTRODUCTION
disproportionates to regenerate
1
and 0.5 equiv
azobenzene. Previous computational work by Wang10
confirmed the feasibility of this overall mechanistic
picture through DFT calculations at the M06-L/6-
311++G(d,p)/PCM level of theory with geometries
obtained in the gas phase at the B3LYP/6-31G(d,p) level of
theory. Wang suggested that the mechanism proceeds
through an unconventional elimination pathway that
requires azobenzene coordination to release pyrrole,
ostensibly avoiding discrete formally TiII intermediates.
Ti is an ideal metal for sustainable catalysis—earth
abundant, environmentally benign and easily recycled.1-2
Considering these properties, Ti catalysis is underutilized
in organic method development compared to the earth
abundant first row late transition metals (Fe, Co, Ni).3-5
Two significant challenges in designing early transition
metal-catalyzed reactions are the oxophilicty of the
metals and the thermodynamic stability of their highest
oxidation states. In fact, Ti-catalyzed reactions involving a
formal oxidation state change at the metal center (i.e.
TiII/TiIV) are quite rare.5
We recently reported a Ti-catalyzed oxidative nitrene
transfer reaction that we speculated proceeds through a
formally TiII/TiIV catalytic cycle: the [2+2+1] coupling of
alkynes with diazenes to yield multisubstituted pyrroles.6
In our initial report, we proposed the mechanism shown
in Figure 1: first, a Ti imido complex 1 undergoes [2+2]
cycloaddition
with
an
alkyne
to
form
an
azatitanacyclobutene 2, identical to the first step of Ti-
catalyzed alkyne hydroamination.7-9 Next, 2 inserts a
second
equivalent
of
alkyne
to
yield
an
azatitanacyclohexadiene 3, which then undergoes
reductive C-N coupling to yield the pyrrole product and a
TiII species 4. 4 could be reoxidized via coordination of
azobenzene to form an η2-hydrazido complex 5, which
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 20
stabilizing solvent interactions also allow for facile pyrrole
dissociation. This mechanistic knowledge has also been
Ph
N
R
10%
py3Cl2Ti(NPh)
1
2
3
4
5
6
7
8
R
R
2
+
0.5
PhN=NPh
extended into a combined computational and physical
organic study of the stereoelectronic effects on the
regioselectivity of unsymmetrical alkyne couplings,
revealing that sterics predominantly control the
selectivity of terminal alkyne reactions, while inductive
effects dominate for internal unsymmetrical alkynes such
as phenylpropyne.
110 oC, PhCF3
R
R = H, TMS, Alkyl, Aryl
R
R
R
0.5 PhN=NPh
diazene
disproportionation
LnTiIV NPh
[2+2]
cycloaddition
1
R
Ph
NPh
LnTiIV
Ph
N
9
2. METHODS
N
R
NPh
R
LnTiIV
5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.1 Experimental Methods. All air- and moisture-
sensitive compounds were manipulated in a glovebox
under a nitrogen atmosphere. Solvents for air- and
moisture-sensitive reactions were vacuum transferred
from sodium benzophenone ketyl (C6D6) or CaH2 (CDCl3,
C6D12, C7D8) or pre-dried on a Vacuum Atmospheres
Solvent Purification System (hexanes, toluene, benzene,
THF, Et2O, DCM, and TFT). Solvents were stored over
CaH2 and filtered through dry basic alumina before use.
Azobenzene was purchased from TCI America (100 g).
Commercial azobenzene is typically contaminated with 0-
5% methanol and nitrobenzene; flash chromatography
using hexanes and grinding the isolated product in a
mortar and pestle before drying in-vacuo yields
analytically pure azobenzene. C6D5Br was prepared
following literature procedure.14 [py2TiCl2(NtBu)]2 and
py3TiCl2(NtBu) were prepared according to literature
procedure.15 All liquid alkynes were freeze pump-thawed
three times, brought into the glove box and passed
through activated basic alumina before being stored at -35
°C. 1H, 13C, 15N, HMBC, HSQC, NOE and No-D NMR
spectra were recorded on Varian INOVA 500 MHz, Bruker
Avance III 500 MHz, Bruker Avance III HD 500 MHz, or
Bruker Avance 400 MHz spectrometers. Chemical shifts
are reported with respect to residual protio-solvent
impurity for 1H (s, 7.16 ppm for C6D5H; s, 7.27 for ppm
CHCl3), solvent carbons for 13C (t, 128.39 ppm for C6D6; t,
77.16 ppm for CDCl3), referenced to tetrachloroethane (s,
5.53 ppm) or trimethoxybenzene (s, 6.02 ppm) in C6D5Br.
Full experimental details and kinetic traces are available
in the Supporting Information.
2
R
R
R
R
R
PhN=NPh
“TiII”
Ph
N
R
R
4
insertion
LnTiIV
R
reductive
elimination
R
R
3
Figure 1. Proposed mechanism of Ti-catalyzed formal [2+2+1]
cycloaddition of alkynes with azobenzene.
We are interested in developing a deeper experimental
and theoretical understanding of the mechanism of this
reaction in order to design new Ti-catalyzed nitrene
transfer reactions, and further understanding the
requirements for catalytic turnover in low-valent early
transition metal-catalyzed processes. In particular, we are
interested in understanding (1) the structure of the active
species in catalysis and the role of catalyst structure in
reactivity; (2) the kinetic interplay between the two steps
involving alkyne cycloaddition and insertion; (3) how and
why C-N reductive elimination occurs; and (4) the role of
azobenzene in C-N reductive elimination and catalyst
reoxidation.
Herein, we report a detailed kinetic and computational
study on the Ti-catalyzed formal [2+2+1] pyrrole
synthesis. The main kinetic manifold of the reaction
involves two equilibria, a Ti imido monomer/dimer
equilibrium followed by
a
[2+2] alkyne/Ti imido
cycloaddition equilibrium. Geometry optimization of
intermediates in solvent turns out to be critical to
uncovering several important details of the catalytic cycle.
Additionally, we perform an exhaustive DFT analysis to
determine which ligands are active during the catalytic
cycle and find that coordinatively unsaturated complexes
have the lowest transition state barriers. By using these
geometries and a combination of Intrinsically Bonding
Orbitals (IBOs)11-12 and Natural Bond Orbitals (NBOs),13
we show that this reaction proceeds through a true
reductive elimination pathway, where discrete “TiII”-like
species are present. NBOs have previously been used
successfully in Ti compounds to describe bonding in Ti-
imido complexes.7b This reductive elimination occurs via
a surprising electrocyclic pathway similar to the Nazarov
cyclization. The resulting TiII pyrrole species is stabilized
through backbonding into the pyrrole π* system, but
2.2 Computational methods Geometry optimizations
were performed using the Gaussian 09 program version
d01.16 Unless otherwise noted, all calculations were
formed using the M06 functional,17 the 6-311G(d,p) basis
set,18 the ultrafine grid, and the SMD solvation model19 in
the experimentally used solvent PhCF3 (ε = 9.18).
Calculations performed for symmetric alkynes used the
superfine grid. All geometries were characterized by
frequency analysis calculations to be minima (without
imaginary frequency) or transition states (having only one
imaginary frequency). The refined energies were
corrected to free energies at 383.15K and 1 atm. Intrinsic
bond orbital (IBO) calculations were performed in
MOLPRO 2015.120 using M06/def2-TZVP.21 Density fitting
was employed to accelerate calculations in MOLPRO
ACS Paragon Plus Environment

Page 3 of 20
Journal of the American Chemical Society
using the def2-TZVP JK-fitting basis set. Molecular and
1
2
3
4
orbital depictions were made using the IBOView
program.11-12 For NBO calculations the NBO 5.G13 program
was used and the d orbital occupation numbers were
generated following the protocol in ref. 22.
5
6
7
8
3. RESULTS AND DISCUSSION
3.1 Experimental Studies and Kinetic Analysis.
First, we set out to determine the kinetic robustness of
the reaction of 4-octyne with azobenzene catalyzed by
[py2TiCl2NtBu]2 by in-situ 1H NMR same-excess analysis.23-
9
24
The dimeric catalyst [py2TiCl2NtBu]2 was chosen due to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
greater solubility over previously-reported py3TiCl2(NR)
species, and because formation of NtBu pyrrole products
then provides information on the amount of catalyst
activation. In experiment 1, 0.44 mmol azobenzene and
2.3 mmol 4-octyne were reacted with 0.04 mmol
[py2TiCl2NtBu]2. In experiment 2, 0.34 mmol azobenzene
and 1.77 mmol 4-octyne were reacted with 0.04 mmol
[py2TiCl2NtBu]2. Overlay of these two reactions produces
good overlapping kinetic traces indicating no catalyst
decomposition and no significant catalyst induction
period caused by the NtBu group (Figure 2).
.
Figure 2. Same-excess experiments of [py2TiCl2NtBu]2-catalyzed
pyrrole formation from 4-octyne and azobenzene indicates no
catalyst decomposition or product inhibition. Reaction Conditions:
Exp 1: 0.44 mmol PhNNPh and 2.3 mmol 4-octyne were reacted with
0.04 mmol [py2TiCl2NtBu]2 in 0.5 mL C6D5Br at 115 °C. In Exp 2, 0.34
mmol PhNNPh and 1.77 mmol 4-octyne were reacted with 0.04
mmol [py2TiCl2NtBu]2 in 0.5 mL C6D5Br at 115 °C.
Having established that this catalyst system is well-
behaved, we next sought to determine the rate law using
Variable Time Normalization Analysis (VTNA) of each
individual component of the reaction.27-28 In VTNA
experiments, the time scale is normalized as a time
integral (approximated via the trapezoid rule) of one
component of the reaction to remove the kinetic effect of
that component from the reaction profile.27 By
systematically varying concentrations of each component,
one
can
rapidly
construct
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Best-fit rate law determination of [py2TiCl2NtBu]2-catalyzed pyrrole formation from 3-hexyne and azobenzene via variable time
normalization analysis (VTNA). Full reaction conditions can be found in Table S1.
graphical overlays that indicate component order. The
VTNA overlay plots of each component of the reaction of
3-hexyne with azobenzene catalyzed by [py2TiCl2NtBu]2
(catalyst, 3-hexyne, azobenzene, pyridine) are presented
in Figure 3, and the approximate derived rate law is
provided in eq. 1.
monomer/dimer equilibrium prior to reversible [2+2]
imido+alkyne cycloaddition followed by irreversible, rate
determining 3-hexyne 2nd insertion or reductive
elimination. (Figure 4). The specific step(s) at which
pyridine and azobenzene inhibits catalysis cannot be
experimentally determined, since the inhibition could
potentially occur at any Ti intermediate prior to the rate-
determining step.
ꢁ
ꢂꢃ.ꢄ
ꢁ
ꢂ2
Ti
3-hexyne
ꢂꢁ
rate = ꢀ
(1)
ꢁ
ꢂ
PhNNPh pyridine
The half-order dependence on [Ti] indicates
a
monomer-dimer equilibrium and that at this
concentration the bulk of catalyst is dimerized.25-26 This is
consistent with many other Ti-catalyzed reactions
involving Ti imido intermediates such as such as alkyne
hydroamination, wherein bridged imidos are often off-
cycle resting states.29-34 Unlike hydroamination reactions,
however, the Ti dimer species is on cycle in [2+2+1]
catalysis, as azobenzene cleavage requires dimerization
(vide infra). The reaction is first-order inhibited by
[PhNNPh], indicating that productive catalyst reoxidation
by PhNNPh is kinetically unimportant, and instead that
PhNNPh can inhibit catalysis through competitive
binding with alkyne at one or more points along the
catalytic cycle. As a result, the second order dependence
on [3-hexyne] indicates that either second alkyne
insertion or reductive elimination is rate-determining.
Taken together, this rate law indicates a likely Ti
Figure 4. Kinetically relevant steps of Ti-catalyzed 3-hexyne pyrrole
formation proposed from rate law data.
Given the initial mechanistic insights above, we next
turned to computation to more deeply examine the
mechanism. First, we wanted to determine the number
and type of L donor ligands bound to the active species at
each step in the catalytic cycle: the rate law indicates an
inhibition by pyridine and azobenzene, but it is unclear at
ACS Paragon Plus Environment

Page 5 of 20
Journal of the American Chemical Society
what stage this occurs. One could envision pyridine,
alkyne, or azobenzene coordinated to the Lewis acidic
1
2
3
metal during catalysis. Additionally, we were interested in
probing the steps of catalysis that are not kinetically
4
observable
(reductive
elimination
and
diazene
5
6
disproportionation) to determine at a fundamental level
how and why this unique TiII/TiIV mechanism takes place.
7
8
9
3.2 DFT analysis of the mechanism and effects of
Ligands. We first examined the energetics of the
catalytic reaction of 2-butyne with azobenzene to
determine the rate-determining step in the mechanism
and the effects of different ligands on the individual steps.
The precatalyst py3TiCl2(NPh) can potentially undergo
dissociation of pyridines such that 1, 2, or no pyridine
ligands are bound; furthermore, 2-butyne or azobenzene
could also act as a ligand. Therefore, we performed a
series of calculations such that we can isolate what
ligands are most probably bound to Ti for each
intermediate and transition state over the course of the
reaction. In general, there is no significant barrier with
the addition or removal of a pyridine ligand or alkyne;
however, the same cannot be said about the addition of
azobenzene, as we discuss later in this section.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5 shows the free energy profile for the formation of
N-phenyl tetraethylpyrrole from 2-butyne. Dissociation
of one pyridine from py3TiCl2(NPh) (CAT) to IM1 is
favorable by 8.2 kcal/mol in PhCF3, while the removal of
two pyridines is roughly thermoneutral. However,
liberation of all 3 pyridines to generate a naked 3-
coordinate complex, binding of 2-butyne, and binding of
azobenzene are all shown to be unfavorable. Therefore,
at the start of the catalytic cycle Ti is likely to be
coordinated to one or two pyridines. This is consistent
1
with H NMR spectroscopic evidence that shows that the
pyridine trans to the imido is already in coordination
equilibrium at room temperature; thus, at elevated
temperature more complex pyridine coordination
equilibria are likely involved.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 20
1
2
3
4
5
reductive
elimination
pyrrole
release
alkyne
insertion
[2+2]
cycloaddition
6
7
L
Cl
Cl
Ph
TiIV
N
8
9
L
Ph
Cl
TiIV
TS2
N
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cl
N
Cl
Cl
TiIV
L
IM4
TS1
Ph
N
Cl
IM2
TS3
+
L
TiII
IM3
Cl
52.6
47.7
46.2
35.4
29.6
IM7
IM1
41.7
41.5
28.0
24.3
23.4
L
37.5
37.3
28.6
22.7
22.0
IM5
CAT
31.3
29.7
15.9
12.7
8.6
24.3
22.7
10.3
8.2
30.4
22.1
19.8
9.2
IM6
0 kcal/mol
Ph
Ph
Cl
Cl
7.3
6.9
-2.1
-4.0
-4.6
TiIV
Cl
Cl
Ph
N
N
N
N
15.9
11.2
9.1
-0.1
-8.2
TiIV
7.2
6.6
L
py
Cl
py
Cl
Ph
N
Ti
Ph
L
Ph
Cl
Cl
TiIV
py
IM8
N
L
Ti
Ph
N
Cl
Cl
Ph
N
Cl
Cl
TiIV
-4.1
-8.4
-13.0
-25.9
-36.1
L
16.0
5.9
-6.1
-24.3
-41.5
L
Ti
Cl
Cl
0 pyridine
1 pyridine
2 pyridine
2-butyne
-33.3
-35.2
-41.5
-53.4
-55.1
L =
Ph
Ph
N
N
azobenzene
TiII
TiIV
L
Cl
Cl
L
Cl
Cl
Figure 5. Free energy profile for the formation of N-phenyl tetramethylpyrrole from 2-butyne, azobenzene and py3TiCl2(NPh). The reaction
pathway is calculated using L = None (black), 1 pyridine (red), 2 pyridines (blue), 2-butyne (green), and azobenzene (orange). The geometries for
all ligand combinations are provided in the Supporting Information. The free energy profile to close the catalytic cycle from IM8 to CAT’ can be
found in Figure 11.
ACS Paragon Plus Environment

Page 7 of 20
1
Journal of the American Chemical Society
Figure 6 compares direct pyrrole release to form TiII
with azobenzene coordination to dissociate pyrrole,
which would ostensibly avoid a discrete formally TiII
intermediate through direct electron transfer from
pyrrole to azobenzene. Our calculations suggest that the
pathway requiring azobenzene binding has a transition
state barrier of 23.1 kcal/mol compared to only 18.9
kcal/mol for direct pyrrole dissociation. These results are
in direct contrast to previous calculations from the Wang
group, who proposed a “redox neutral” associative
interchange mechanism that was predicated on
azobenzene coordination prior to pyrrole release to avoid
discrete, free TiII intermediates.10 We believe the inclusion
of solvent polarization effects within our geometry
optimizations leads to the differences in conclusions:
other groups have optimized Ti catalysts in implicit
solvation and have found good agreement with
experimental kinetics.35 Additionally, we have found
experimentally that stoichiometric pyrrole production
(and subsequent catalytic alkyne cyclotrimerization by
TiII) is possible in the absence of azobenzene.36 Taken
together, this suggests that azobenzene binding for
pyrrole release is not a requirement for [2+2+1] pyrrole
synthesis, and that free TiII intermediates are plausible
under catalytic conditions.
After pyridine loss, 2-butyne can then coordinate to Ti to
2
3
4
5
6
7
make IM2, which is an endergonic process for all possible
ligand combinations. As the catalyst coordinates to the π-
bond of the alkyne, the lowest energy IM2 structure
contains only one pyridine to minimize steric effects.
This species then reacts through [2+2] cycloaddition to
the imido Ti=N bond through TS1, with a barrier of 13.6
8
kcal/mol (assuming
1
bound pyridine) to form
9
azatitanacyclobutene (IM3). The relative free energy
differences between 0, 1, or 2 pyridines bound to IM3 are
relatively small, suggesting that all of these complexes
may play a role during the cycloaddition step. Likewise,
coordination of the second equivalent of 2-butyne to IM4
can readily occur with either 0 or 1 pyridine.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The 2-butyne insertion transition state TS2 proves to be
the rate-determining step. Unlike previous steps where
several ligand combinations were energetically feasible,
this step has significant free energy differences between
the different ligand combinations, strongly suggesting
that the catalyst has a single pyridine ligand during TS2.
The reaction barrier is 5.3 kcal/mol compared to IM4,
29.6 kcal/mol compared to the starting catalyst while
being exergonic to form azatitanacyclohexadiene IM5 by
33.6 kcal/mol. This also accounts for the observed
inhibition by azobenzene and pyridine, since
coordination of either of these species significantly raises
the energy of all species through TS2.
Next, C-N reductive elimination from IM5 occurs to
form a TiII-bound pyrrole IM6. Remarkably, despite
forming a formally TiII species, the formation of IM6 has a
relatively small barrier of 10.6 kcal/mol for the single
pyridine bound catalyst and is exergonic. We discuss this
step in detail in Section 3.3 to show how it is likely that a
true TiII intermediate is formed at this stage of the
catalytic cycle.
The next step is release of the pyrrole product to form
IM7. Surprisingly, we find that even with a single
pyridine ligand the energy required to dissociate the
pyrrole is only 18.9 kcal/mol, significantly lower than the
rate determining step. This suggests that the formation
of TiII is readily accessible.
Since IM6 with bound azobenzene is significantly lower
in energy than all other ligands, we examined this step in
greater detail. It is clear from Figure 5 that azobenzene
being bound to Ti prior to IM6 is significantly
unfavorable for productive catalysis, consistent with the
kinetic inhibition observed experimentally. Therefore, to
proceed over TS3 to form IM6 the catalyst must have
either 0 or 1 bound pyridine ligands. However, after the
reductive elimination, an azobenzene could bind to IM6
to aid in pyrrole release via associative displacement as
suggested by Wang.10 Therefore, we calculated the barrier
for associative displacement of pyrrole by azobenzene,
shown in Figure 6 as TS4.
Figure 6. Comparing the free energy required to dissociate pyrrole
from TiII to the transition state barrier for azobenzene binding to the
Ti-pyrrole complex. Direct pyrrole dissociation from TiII in IM6 is
the lower energy pathway.
Finally, IM7 is reoxidized by binding azobenzene to
form IM8. IM8 is best described as a TiIV hydrazido2-
species wherein the electron pair from the TiII has been
fully backdonated into the N-N π* of azobenzene (vide
infra), as shown in Figure S114c. Overall, the catalytic
production of N-phenyl tetraethylpyrrole is exergonic by
55.1 kcal/mol with a barrier of 29.6 kcal/mol if the most
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 20
energetically favorable ligands are used, which is a
complex. But the question remains if formal reduction
1
2
3
4
5
6
7
8
occurs during C-N bond formation, or if the pyrrole acts
as a fully redox active ligand through bilateral donation
and back-donation, obviating the need for charge
distribution onto Ti. Therefore, we calculated the Natural
Bond Orbitals (NBO) and performed Intrinsic Bond
Orbitals (IBOs) analysis to gain insight into this step.
reasonable barrier for the 110 °C temperature.
In order to further validate the computationally-derived
energy values, we carried out Arrhenius analysis on the
reaction of 4-octyne with azobenzene catalyzed by
[py2TiCl2NtBu]2. Figure 7 shows the Arrhenius-like plot for
4-octyne giving an activation enthalpy of 16.3 ± 0.6
kcal/mol. Computationally, the activation enthalpy from
CAT to TS2 for 4-octyne is 21.1 kcal/mol; therefore, this
represents relative good agreement between our DFT
model and the experimental result.
In complicated binding situations involving noninnocent
ligands and covalent metal-ligand bonding, formal
oxidation state assignment from first principles or DFT
calculations can become ambiguous or arbitrary—after
all, oxidation state is merely a formalism used to describe
extreme ends of a spectrum of bonding. However, it is
often useful to describe systems in terms of relative
oxidation/reduction. D’Acchioli et. al. suggested using the
d-orbital occupations22 based on NBO theory instead of
atomic charge to infer to the natural oxidation state of a
transition metal,39 since simple changes to the primary
covalent coordination sphere (i.e. changing from alkoxide
to amide) can dramatically impact the computed charge
on the metal. Therefore, we use both Natural Bond
Orbitals (NBO) and Intrinsic Bond Orbitals (IBOs) to
argue that C-N bond formation from IM5 to IM6 occurs
through a formal +4/+2 reductive elimination.
The results of 3dz2 NBO orbital occupation calculations
for all pyrrole-forming steps of the catalytic cycle
(assuming 1 bound pyridine) are listed in Table 1, with the
full list of all d-orbitals in Table S8. All of the d orbital
occupations remain mostly constant throughout the
reaction except for the 3dz2 orbital occupation. By
analyzing the 3dz2 orbital and relating the occupation of
this orbital to the formal oxidation state for compounds
with known oxidation states, we can infer the oxidation
state for all intermediates. As a benchmark for TiII
complexes, the 3dz2 values for TiCl2, [TiCl4]2-, TiCl2py2,
and TiCl2py4 were found to range between 0.65 to 0.82. As
a benchmark for TiIV complexes, IM1-IM5 and TiCl4 were
calculated to have 3dz2 values between 0.29 to 0.44.
Therefore, intermediates with 3dz2 values closer to 0.4
can be assigned a formal +4 oxidation state and closer to
0.6 a +2 oxidation state.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The overall rate constant in an Eyring equation37 type
form is:
∗
∗
∆ꢋ ∆ꢎ
ꢀꢇꢈ
ℎ
ꢌ
ꢏ
ꢀ = ꢅ ꢆ
ꢉ ꢊ ꢌ
ꢊ
ꢍ
ꢍ
where ∆ꢐ∗ and ∆ꢑ∗ is the change in enthalpy and entropy,
from CAT to TS2, ꢅ is the transmission coefficient, ℎ is
Plank’s constant, ꢀꢇ is Boltzmann’s constant. By using the
experimentally-measured rates and our DFT calculations
for the enthalpy (21.1 kcal/mol) and entropy (18.8 cal/mol
K), we can then extract that ꢅ is 0.017. The small
transmission coefficient shows that this reaction occurs in
the high-friction regime of Kramer theory,38 showing that
solvent viscosity likely plays a role in the kinetics. There is
a large geometry change that occurs from TS2 to IM5, yet
a
small geometry change between IM4 and TS2;
therefore, there is a relatively large solvent reorganization
required after the transition state to proceed to products,
which leads to a small transmission coefficient.
From this analysis, we assign IM5 a formal +4 oxidation
state (3dz2 = 0.44) and IM6 a formal +2 oxidation state
(3dz2 = 0.61). However, the 3dz2 occupation of 0.61 is on
the lower end of occupation for TiII, which does suggest
strong bilateral donation and back-donation into the
pyrrole π*. This backdonation can be seen in the
calculated bond lengths of the pyrrole in IM6, where C1-
N, C4-N, and C4-C3 are elongated (Figure 8). Similarly,
the 3dz2 value (0.64) for TiCl2py4, which can also
backdonate into the ligand framework, is also on the
lower end of “TiII” occupation. Interestingly, TS3, the C-N
bond-forming step, has a 3dz2 of 0.54 suggesting that the
2-electron reduction is happening through this transition
state. IM7 is assigned a +2 oxidation state before the
Figure 7. Arrhenius-like plot of the reaction of 4-octyne with
azobenzene catalyzed by [py2TiCl2NtBu]2. Reaction conditions can be
found in Table S7.
3.3 Mechanism and NBO/IBO analysis of C-N
reductive elimination. Our calculations show that
dissociation of pyrrole from TiII intermediates is more
favorable than azobenzene binding to the TiII-pyrrole
ACS Paragon Plus Environment

Page 9 of 20
Journal of the American Chemical Society
catalysts binds with azobenzene, after which it returns to
the +4 oxidation state in IM8.
A
New NꢀC4 ! ꢀbond
1
2
3
4
5
6
Table 1. The occupation number of the 3dz2 orbital compared with
the formal oxidation state of the Ti atom in the complex. Species
labels match with structures in Figure 5. This analysis suggests that
complexes with an occupation number closer to 0.4 should be
assigned the formal +4 state, and complexes with an occupation
number closer to 0.6 should be assigned the form +2 state.
N lone pair/
TiꢀN πꢀbond
B
7
8
New
backbonding
π* for Ti
Species
Occupation numbers
(3dz2 orbital)
Formal Oxidation State
TiꢀC
!
ꢀbond
9
TiꢀN ! ꢀbond
IM1
0.31
0.30
0.35
0.38
0.37
0.40
0.44
0.54
0.61
0.72
0.38
0.82
0.77
0.68
0.65
0.29
+4
+4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IM2
TS1
+4
C
IM3
+4
NꢀC4: 2.638
TiꢀN: 1.998
TiꢀC4: 1.979
TiꢀN: 2.122
TiꢀC4: 2.078
1.949
Cl
Ti
py
Cl
IM4
+4
C4
C4
Ti
1.465
1.959 C4
1.441
C3
1.365
1.395
C3
1.943
Ph
Ph
N
Ph
N
N
C3
1.441
TS2
+4
Ti
1.398
1.448
1.430
1.399
1.371
C1 C2
1.405
C1 C2
1.391
C1 C2
1.380
IM5
+4
IM5
TS3
IM6
TS3
+2/+4
+2
Figure 8. IBOs of the reactant (L, IM5), transition state (M, TS3) and
product (R, IM6) in the reductive elimination step. IBOs show (A)
the lone pair on the N atom/Ti-N π-bond in IM5 (blue orbital)
rotates and attacks C4 to form the new C-N σ bond in an
electrocyclic fashion; (B) orbital rotation for π backbonding with Ti;
and (C) an arrow pushing diagram of the reductive elimination along
with calculated bond lengths (in Å).
IM6
IM7
+2
IM8
+4
TiCl2
[TiCl4]2-
TiCl2py2
TiCl2py4
TiCl4
+2
Figure 6 shows the IBOs for the reductive elimination
step from IM5 to IM6. Figure 6a shows that the lone pair
on the N atom in IM5 (blue orbital) rotates and attacks
C4 to form the new C-N σ bond in an electrocyclic
fashion, similar to how a C-C bond is formed in a Nazarov
cyclization of divinylketones.40 Simultaneously the orange
orbital (π bond on C1 and C2) and the magenta orbital (π
bond on C3 and C4) simply shift during the reductive
elimination and become localized on C1, C2, and C3.
Concurrent with electrocyclization, Figure 6b shows the
fate of the Ti-C and Ti-N σ-bonding orbitals. As the
electrocyclization takes place, these two orbitals rotate
perpendicular to the forming pyrrole ring, becoming the
basis for a π backbonding interaction with Ti. The net
result of this electron movement (Figure 6c) is to generate
a bound pyrrole whose binding can be described as an η3-
allyl interaction between C1, C2, C3 and Ti, and a strong π
backbonding interaction between T, N, and C4. This
bonding picture is consistent with the calculated bond
lengths in IM6, where C1-C2 and C2-C3 are average (~1.4
Å), while C3-C4, C4-N, and N-C1 are elongated (~1.45 Å)
compared to normal pyrrole bond lengths.
+2
+2
+2
+4
Further evidence for the reductive elimination
mechanism is obtained through intrinsic bond orbital
(IBO) analysis. By obtaining IBOs along the reaction
coordinate, the continuous rearrangement of IBOs can be
correlated to electron flows and bond rearrangements,
and then expressed by the curved arrow formalism.11,12 By
construction, a minimal number of IBOs change their
a reaction, to make it clear and
straightforward to analyze the mechanism and identify
electron flow.
nature during
This electrocyclic reductive elimination is unique, and
provides an important piece of data on C-X (X =/= H)
reductive eliminations on Ti: direct M-C σ reductive
coupling on Ti is virtually unknown; it appears as though
in almost all cases Ti organometallics would energetically
ACS Paragon Plus Environment

Journal of the American Chemical Society
abstraction,41-42 radical decomposition,43
C2H4Cl2, C6D5Br, and PhCF3 all have similar rates.
Page 10 of 20
prefer
H
substitution,44 or electrocyclization to classical 2-electron
reductive elimination pathways seen with late transition
metals. In fact, in our previous report on Ti-catalyzed
alkyne/alkene/nitrene coupling, direct C-N bond forming
was never observed.45 Since these systems had semi-
saturated ring systems, they are unable to undergo
electrocyclization and instead access alternative reductive
pathways. Similarly, the Ti-catalyzed Pauson-Khand
reaction, an example of apparent C-C direct reductive
coupling, may instead proceed through a related π-
electrocyclization-type pathway.46-47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.4 Solvent effects on the mechanism and rate. Next,
the effect of solvents on the rate of catalysis was examined
by varying the solvent dielectric from 2.03 (C6D12) to 10.5
(C2H4Cl2), and the data summarized in Figure 9. There is
a strong correlation between rate and dielectric constant,
with high dielectric solvents (C2H4Cl2, C6D5Br, PhCF3)
yielding fast, productive catalysis and low dielectric
solvents (cyclohexane, C6D6, C6F6) performing poorly.
Additionally, there is no obvious arene coordination or
other solvent coordination effect to the rate since
Figure 9. Solvent screen of [py2TiCl2NtBu]2-catalyzed pyrrole
formation from 3-hexyne and azobenzene. Reaction conditions can
be found in the Figure S34.
alkyne
insertion
reductive
elimination
pyrrole
release
L
Cl
Cl
C6H5CF3
C6H5Br
C6F6
Ph
calculations on a subset of these solvents to probe the
effect of dielectric constant on the rate.
TiIV
N
Figure 10 shows the calculated catalytic cycle for 3 of
the experimentally used solvents, PhCF3, C6D5Br and C6F6.
Overall, the catalytic picture looks largely the same, with
TS2 being the highest overall energy of the catalytic cycle.
However, contrary to experimental expectations, the free
energy difference between CAT and TS2 actually increases
with increasing dielectric constant with barriers of 29.6,
29.1, and 25.5 kcal/mol barriers for PhCF3, C6D5Br and
C6F6, respectively. Since the TS2 barrier controls the rate,
this data suggests that the increase in the barrier with
high dielectric constant is contrary to experiment. Given
that we are using an implicit solvation model, the
associated errors with such a model can certainly lead to
3-5 kcal/mol errors.48 Furthermore, as discussed in the
Supporting Information, we only changed the dielectric
constant of the solvent and ignored other changes in the
solvent such as the thermal expansion coefficient or the
surface tension at interface. Additionally, we showed that
this reaction takes place in the high friction regime of
Kramer rate theory; therefore, the solvent viscosity, which
was not taken into effect here, plays a role in the kinetics.
Therefore, the discrepancy between our calculations and
experiments could be from not including these effects.
TS2
Ph
Ph
N
N
Cl
TiII
Cl
Cl
Cl
+
TiIV
L
29.6
29.1
25.5
L
IM7
TS3
8.7
8.4
5.9
IM5
6.6
5.4
2.3
IM6
-4.0
-4.0
-8.2
Ph
Ph
Cl
Cl
N
N
-13.0
-13.5
-17.1
TiIV
Ph
N
TiIV
L
Cl
Cl
-50.5
-53.4
-57.3
Ph
Cl
Ph
Cl
N
N
L
TiII
TiIV
L
L
Cl
Cl
IM8
Figure 10. The free energy profile for the alkyne insertion, reductive
elimination, and pyrrole release step in PhCF3 (black), C6F6 (red), and
C6D5Br (blue).
Curiously, the data shows that PhCF3, C2H4Cl2, and
C6D5Br all have very similar rates, despite the fact that
C6D5Br’s dielectric constant is only 5.4, while PhCF3 and
C2H4Cl2 are 9.18 and 10.5, respectively. This step function
behavior in the rate suggests a deeper mechanistic feature
in the role of the solvent; therefore, we performed
Despite this, an interesting trend can be seen which
might suggest why experimentally the reaction rate
increases with dielectric constant. The free energy
difference between IM6 and IM7 is increasing with
ACS Paragon Plus Environment

Page 11 of 20
Journal of the American Chemical Society
decreasing dielectric constant, going from 18.9 (PhCF3) to
21.9 (C6D5Br) to 25.8 kcal/mol for C6F6. This data shows
1
2
3
4
5
that high dielectric constant solvents are able to more
adequately stabilize free TiII species. Therefore, it could
be that in C6F6 slower ligand dissociation rates at various
steps of catalysis may play a role in kinetically relevant
equilibria or related processes.
6
7
8
9
Additionally, it is important to note how sensitive these
calculations are to fully relaxing the geometries in the
associated solvent. We also performed single-point
energy calculations in these three solvents using PhCF3
solvent-optimized geometries (Figure S111). Using this
simplified protocol, there is a dramatically higher energy
difference between IM6 and IM7, which would lead to
the inaccurate conclusion that the free energy difference
between IM6 and IM7 was becoming the rate-
determining step for C6F6. This further demonstrates the
large role solvent plays in the dissociation of pyrrole.
Additionally, the differences between Figure 10 and Figure
S111 emphasize the need for careful computations in the
description of solvent effects.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 11. Free energy profile for the regeneration of the Ti imido catalyst via disproportionation.
forming an unsymmetric bis imido-bridged four-
membered ring IM10 where the non-reacted azobenzene
remains coordinated to one Ti. This dimerization
pathway is consistent with previous crossover studies
which indicated no azobenzene N-N scrambling during
catalysis.6 Subsequently, there is a reaction barrier of 9.6
kcal/mol to liberate this azobenzene to generate IM11, a
bridging imido dimer. These species are often the resting
states in other Ti imido-catalyzed reactions such as
hydroamination,29-34 and in fact this species is close in
energy to CAT’.
3.5 Mechanism for regeneration of the catalyst via
disproportionation. Next, we computed the
regeneration of the active Ti imido (CAT’) from the η2-
hydrazido adduct IM8. Previously, we hypothesized that
this reaction proceed through a dimerization process to
make a bridged Ti2N4 6-membered metallacycle that
would then eject azobenzene and regenerate 2 Ti imidos
in a retro-[2+2+2] reaction. However, we were unable to
experimentally interrogate this mechanism because
simple ligand loss from model complexes was rate-
limiting.
There are multiple energetically accessible pathways
from IM11 to CAT’, we show the lower energy pathway
here and a secondary pathway in the Supporting
Information (Figure S112), which uses different
regoisomers. The pathway shown here involves first
gearing the imido-bridge dimer IM11 to a chloro-bridged
dimer IM13 (the experimental crystal structures of
[TiCl2(NC6H4Me)py2]2 and [TiCl2(NC6F5)py2]2 are chloro-
bridged29b) in contrast direct splitting of IM11 as proposed
by the Wang group.10 In this chloro-bridged splitting
pathway, IM11 first isomerizes by overcoming a 17.4
kcal/mol barrier in TS7 to move from a bis imido-bridge
to a mixed imido/chloro-bridge IM12. Next, the second
isomerization takes place, this time only requiring 9.5
kcal/mol of energy through TS8 to form bis chloro-
The computed mechanism for regeneration of the Ti
imido catalyst is shown in Figure 11. We were unable to
locate a reaction pathway that proceeded through the
originally proposed retro-[2+2+2].6 Here, we start with
two IM8 complexes with one pyridine on each complex
since pyrrole release likely occurs with one pyridine (vide
supra). For convenience, we reset the energy scale to zero
here, but IM8 is actually -53.4 kcal/mol relative to the
starting catalyst; therefore, we do not expect, nor is
experimentally observed, that any step in the regeneration
pathway to be rate determining. Two equivalents of IM8
first dimerize through one chlorine atom and one
azobenzene to form IM9. Next, IM9 then reacts to break
the bridging azobenzene N-N bonds through TS5,
ACS Paragon Plus Environment

Page 13 of 20
Journal of the American Chemical Society
a Reaction conditions can be found in the Supporting Information. Et
bridged IM13. At this point, barrierless addition of
pyridines forms a coordinatively saturated IM14, which is
exergonic by 9.9 kcal/mol. Finally, the dimer IM14
= S43, Pr = S64. bSteric A value;50 cTaft steric parameter;49 dSteric
1
2
3
value;49 eM06/6-311g(d,p)/SMD/ultrafine grid in C6H5Br. Δꢒ≠ TS2 is
the energy difference between CAT and TS2.
breaks through a small 3.3 kcal/mol barrier to regenerate
4
5
6
7
8
9
2 equivalents of CAT’.
Unsymmetrical alkynes introduce
complexity: there are 3 possible regioisomeric products
that result from 2 azatitanacyclobutene intermediates
a new layer of
Regardless of how the dimer dissociates, the free energy
difference between IM11 and CAT’ is relatively small,
suggesting an equilibrium of the dimer and monomer
under catalytic conditions. This matches the
experimentally observed rate order of [Ti]0.5.
(IM3) and
4
azatitanacyclohexadiene intermediates
(IM5). The selectivites for several unsymmetrical alkynes
are shown in Table 3.6 Given these results, we were
interested in determining the exact pathways through
which products were formed, since the 2,4-substituted
head-to-tail coupled product H could be formed through
multiple routes (Figure 13, vide infra). Based on our initial
computational and kinetic results, formation of IM3 is in
most cases reversible, while IM5 is irreversible. Thus, in
order to understand the origin of selectivity, the
equilibrium constants and each individual 2nd insertion
rate must be known. As such, we embarked on full kinetic
and computational analysis of the selectivity manifold of
several unsymmetrical alkynes.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.6 Alkyne Effects on Reaction Kinetics and
Regioselectivity. Next, the effect of alkyne substitution
on the reaction kinetics and product regioselectivity was
explored. First, we experimentally examined the rates of
3-hexyne vs 4-octyne (Table 2) to explore any steric
differences on reaction rate, and compared these to the
computationally-derived relative rates of 2-butyne, 3-
hexyne, and 4-octyne. 2-butyne was not tested
experimentally because its low boiling point precludes
precise concentration measurements at the elevated
reaction temperature. It was found that there was a small
observable experimental rate difference between 3-hexyne
Table 3. Product distribution of pyrrole formation with various
unsymmetric alkynes. Taken from ref. 6.
n
and 4-octyne. Although Et and Pr groups have similar A-
values, their ligand repulsion energies (Er) and steric
values (Es) are quite different, which appears to have
some small influence on rate-determining 2nd insertion
(TS2).49 Consistent with this observation, there is only a
small computed energy difference between TS2 for the 3
internal alkyne substrates. The experimental kobs
difference between 3-hexyne and 4-octyne correlates to a
free energy difference of 1.4 kcal/mol, which is in excellent
agreement with the computational result of 1.4 kcal/mol.
Further computational analysis of the complete catalytic
cycle also indicates only small changes in energies
between 2-butyne, 3-hexyne, and 4-octyne for other
intermediates and transition states (Figure S113).
R1,R2
H, nBu
H, tBu
H, Ph
0.13
0
1
1
1
1
0.13
0
0
0.25
0.91
Me, Ph
0.45
Table 2. Alkyne substituent effects on the rate of [2+2+1] pyrrole
synthesis.
For the purposes of experimentally examining the kinetics
of unsymmetrical reactions, phenylpropyne (PhCCMe)
derivatives were selected. PhCCMe was determined as the
best candidate for kinetic analysis because terminal
alkynes are typically plagued by alkyne trimerization side
reactions.6 Again we turned to VTNA to determine the
rate equation of the reaction of PhCCMe with azobenzene
and [py2TiCl2NtBu]2. The VTNA reactions (Figure 12) give a
rate equation consistent with the computational analysis
(eq. 2). In this case, [PhCCMe] is still second order at low
concentrations and both [PhNNPh] and [pyridine] were
still found to give first order inhibition of the reaction.
[Ti] is still 0.5 order, indicating that a change in the
alkyne substrate has, unsurprisingly, little effect on the
catalyst monomer-dimer equilibrium. However, at high
[PhCCMe] the order changes to [PhCCMe]1.5 (Figure S40).
This change in the rate order indicates that at high
c
ES
d
R =
Kobs x 105
Ab
1.7
ER
Δꢒꢓ TS2
(Calc)e
Me
Et
nPr
-
0
17
29.1
27.1
28.5
5.97 ±0.4
2.1
2.1
0.08
0.31
34
36
0.980
±0.07
Et vs nPr ΔΔꢒꢓ from
Kobs : 1.4 kcal/mol
Et vs nPr ΔΔꢒꢓ
from calcs: 1.4
kcal/mol
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 20
ꢁ
ꢂꢁ
ꢂ1.5-2
Ti PhCCMe
concentration of PhCCMe there is population of [2+2]
rate = ꢀ
(2)
1
2
3
4
cycloadduct IM3 in equilibrium with IM1. Consequently,
this perturbs the rate law from rigorous second order.
This is in contrast to 3-hexyne, which under all tested
conditions is second order.
ꢁ
ꢂꢁ ꢂ
PhNNPh pyridine
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 12. Best-fit rate law determination of [py2TiCl2NtBu]2-catalyzed pyrrole formation from phenylpropyne and azobenzene via variable time
normalization analysis (VTNA). Reaction conditions can be found in Table S3.
Next, the catalytic cycle from precatalyst through IM5
(immediately after irreversible, rate determining 2nd
alkyne insertion) was calculated for 3 unsymmetric
(TS2BE) insertion from IM3B. As a result, all possible
pathways are kinetically accessible, leading to
a
distribution of G, H, and I that agrees well with
experimental data. Product H is the major product
because the lowest energy pathway (IM3A to IM5D) and
the second lowest energy pathway (IM3B to IM5E) both
lead to H.
t
terminal alkynes, nBuCCH, BuCCH, and PhCCH. The
energies of all selectivity-relevant species are presented in
the selectivity manifold in Figure 13. Full computational
details and selectivity determination for these three
alkynes are available in the Table S9.
tBuCCH regioselectivity is dominated by sterics. Both
[2+2] cycloaddition pathways are energetically accessible
n
For BuCCH, there is very little preference for forming
>100 °C, but there is a huge kinetic preference (ΔΔꢒꢓ
=
either [2+2] regioisomer IM3A (electronically favored) or
IM3B, (sterically favored). Despite the fact that it may
seem counterintuitive that Ti would be the sterically
“small” fragment of the Ti=NPh unit, the 2 Cl ligands
provide little spatial hindrance to [2+2] cycloaddition in
TS1 in comparison to the Ph on the imido. In the alkyne
insertion step, TS2, going from IM3 to IM5, there is a
small (ΔΔꢒꢓ = 1.4 kcal/mol) preference for 1,2 insertion
from IM3A (TS2AD, going to IM5D), and little preference
(ΔΔꢒꢓ = 0.1 kcal/mol) for either 1,2- (TS2BF) or 2,1-
7.2 kcal/mol) as well as a thermodynamic preference
(Δꢒ = 4.5 kcal/mol) for formation of the sterically favored
t
n
IM3B due to the increased size of Bu vs. Bu.50 In the 2nd
insertion step from favored IM3B, there is a large steric
preference for 2,1-insertion over 1,2-insertion (ΔΔꢒꢓ
=
6.1 kcal/mol for TS2BE over TS2BF), and as a result tBuCCH
produces solely H.
Finally, PhCCH again has
a
slight kinetic and
thermodynamic preference for the formation of the
ACS Paragon Plus Environment

Page 15 of 20
Journal of the American Chemical Society
product Q is made in lowest quantity since it can only
sterically-preferred IM3B over IM3A. Additionally, there is
1
2
3
4
5
6
7
8
a low-energy pathway for sterically-preferred 2,1-insertion
in the second step (TS2BE) leading to IM5E and ultimately
onto H, which is the experimentally-determined major
product. However, the computational selectivity
significantly favors formation of I. This is because TS2BF
converges to a lowest-energy state that has an arene π-
stacking interaction. Since our simplistic rate theory does
not take ensemble averages into account, the contribution
of other, non-stacked close-in-energy states to the
ensemble partition function at high temperature is
missed, disproportionally increase the rate of forming
IM5F that yields product I. Nonetheless, excluding this
pathway the computational pathways and selectivities for
all three alkynes match experimental data well.
originate from IM3K. Interestingly, much like with
nBuCCH, there is again little kinetic preference for the
regioselectivity of 2nd insertion, leading to significant
quantities of both R and S.
In summary, the regioselectivity of pyrrole formation in
terminal alkynes is driven primarily by sterics, where the
degree of selectivity is dependent on the size and nature
of the alkyne functional group. In contrast, for an internal
alkyne such as phenylpropyne where the steric bias is not
as obvious, inductive effects have a larger impact on
product distribution.51 Ironically, the head-to-tail coupled
product (H or R) is the major product no matter whether
sterics or electronics is the controlling selectivity factor.
This unfortunate phenomenon is a function of the
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
bifurcating selectivity manifold wherein
2 different
pathways can later merge to give a common product.
Figure 13. Reaction pathways leading to 3 regioisomeric pyrroles
from unsymmetrical terminal alkynes. Ln = pyCl2 for all calculated
structures. Full details are available in Table S9. Calculated transition
state and intermediate energies are presented below each structure.
All energies are in kcal/mol and reported with respect to
py3TiCl2(NPh) = 0.0 kcal/mol. Product distributions are standardized
against H. Experimental ratios were taken from ref 6.
Figure 14. Reaction pathways leading to 3 regioisomeric pyrroles
from phenylpropyne alkynes. Ln = pyCl2 for all calculated structures.
Full details are available in Table S9. Calculated transition state and
intermediate energies are presented below each structure. All
energies are in kcal/mol and reported with respect to py3TiCl2(NPh)
= 0.0 kcal/mol. Product distributions are standardized against R.
Experimental ratios were taken from ref 6.
Next, we computationally investigated the regioselectivity
of reactions with PhCCMe (Figure 14). Unlike in the
terminal alkynes, [2+2] cycloaddition is most influenced
by electronic effects of the alkyne substituents:
electronically favored IM3L is both kinetically and
thermodynamically favored over IM3K, and as a result
3.7 Phenylpropyne Substitution Effects and Hammett
Analysis. Given the computational result of the
importance
of
inductive
effects
in
PhCCMe
regioselectivity, we next carried out Hammett analysis
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 20
using various para-substituted phenylpropyne derivatives in TS1K (Figure 16). It is difficult to interpret the effects
1
2
3
4
5
6
7
8
to determine the transition state electronic effects on the
rate of reactivity and regioselectivity. A survey of 5
different phenylpropynes indicates a very strong linear
correlation between rate and electron donicity (Figure 15).
This correlation most likely stems from the increased
Lewis basicity of the alkyne, which leads to an increased
rate of coordination to Ti, which should assist in both
[2+2] cycloaddition and 1,2-alkyne insertion.52
this may have on the insertion or reductive elimination
steps, since the electronic effect now influences the
partial charges on both the reactive intermediate (IM3K or
IM3L) as well as the alkyne.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 16. Partial charge explanation for donor effects in
arylpropyne regioselectivity.
3.8 Catalyst Effects on Reaction Kinetics and
Regioselectivity. Next, we investigated simple anionic
ligand changes on the catalyst using Cl, Br, and I to probe
electronic and steric effects based on ligand donor
parameters (LDP) previously generated for early
transition metals.53-54 These values allow direct
comparison between different ligand sets based on their
donor-ability (LDP) and steric contributions (%Vol) to
the metal center. A general periodic trend can be seen
(Table 4): the larger, poorer donor ligand I is significantly
faster than the smaller, stronger donor ligand Cl.
Interestingly, the rate effect of changing the X ligand is
significantly smaller than simply using a bis(pyridine)
catalyst [py2Cl2TiNtBu]2 instead of py3TiCl2NtBu. Since the
data is only three points, we chose not to over interpret
this data and assign a fitted model to determine the
relative contribution of LDP vs. %BurV. 55
X =
NMe2
OMe
H
0.77
0.62
0.45
0.49
0.39
1
1
1
1
1
0.91
1.13
0.91
1.25
1.01
F
CF3
Figure 15. Top: Hammett plot of various para-substituted
phenylpropynes. Bottom: product distribution of
regioisomers as a function of p-substitution. All product
ratios are standardized against R. Reaction conditions are
available in Table S5.
Table 4. Ligand donor parameter (LDP) effects on the rate of
py3TiX2NtBu-catalyzed pyrrole formation from 3-hexyne and
azobenzene. Reaction conditions can be found in Table S6.
Analysis of the regioselectivity in electronically different
phenylpropynes reveals an interesting trend in
regioselectivity: decreasing electron density on the alkyne
decreases
the
formation
of
3,4-dimethyl-1,2,5-
triphenylpyrrole Q relative to the other two regioisomers
(Figure 15, bottom). We propose that this result indicates
that there is a bias against formation of the electronically
disfavored azatitanacyclobutene intermediate, K. This is
consistent with a picture in which electron-withdrawing
groups on the phenyl ring stabilize partial negative charge
buildup on the carbon adjacent to the phenyl ring in TS1L,
and likewise destabilize positive charge buildup on carbon
Catalyst
Kobs x 105
0.42 ±0.17
0.97 ±0.17
2.34 ±0.11
5.97 ±0.4
LDPa
15.05
15.45
15.80
15.05
%BurVb
16.8
py3TiCl2NtBu
py3TiBr2NtBu
py3TiI2NtBu
18.1
19.2
[py2TiCl2NtBu]2
16.8
aLigand donor parameter.53-54 b% Buried volume from SambVca.56
ACS Paragon Plus Environment

Page 17 of 20
1
Journal of the American Chemical Society
diazenes (H, Me, Et, and i-Pr) yielding the Taft plot
shown in Figure 17 (bottom). Changing from H to Me
causes a large decrease in rate, but further increases yield
only moderate additional slowdown. As was the case in
the azobenzene Hammett, the overall decrease in rate is
most likely due to increased steric bulk at the metal
center inhibiting alkyne coordination events such as [2+2]
cycloaddition and 2nd insertion.
Further analysis of these catalyst effects on
2
phenylpropyne pyrrole regioselectivity yields an increase
in the selectivity for 2,5-dimethyl-1,3,4-triphenylpyrrole S
and decreasing selectivity for Q when moving from Cl to I
(Table 5). We propose a compounding electronic and
steric effect on the reaction mechanism to achieve this
selectivity. We hypothesize that the initial [2+2] is
primarily governed by increased electrophilicity of the
titanium center/polarization of the Ti imido favoring TS1L
and IM3L. This is in agreement with the Hammett results
with substituted arylpropynes. Steric effects then
contribute for the insertion step, where the larger iodide
ligands disfavor putting the sterically encumbered aryl
group α to Ti relative to the smaller chloride ligands.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. X-type ligand effects on product distribution from
phenylpropyne. All product ratios are standardized against R.
Reaction conditions can be found in Table S6.
Catalyst
py3TiCl2NtB
u
py3TiBr2NtB
0.45
0.22
0.14
1
1
1
0.91
1.21
1.41
u
py3TiI2NtBu
3.9 Azobenzene effects on reaction kinetics. Due to
the inverse order in [azobenzene] we hypothesized that
altering the electronics or sterics of the diazene could
potentially lead to a change in rate. Seven p-substituted
aryldiazenes were tested to generate the Hammett plot
shown in Figure 17 (top). A linear correlation was found
between strong EDG (OMe) to weakly EWG (F), giving a
general decrease in reaction rate. We hypothesize this
decrease in rate is due to decreased Lewis acidity of the
metal center containing a more electron rich imido
ligand, which should inhibit alkyne binding. Interestingly,
strongly EWG groups (Cl, OCF3) do not fall within this
trend. Typically large deviations in Hammett plots would
indicate a change in mechanism or a change in the RDS;57-
Figure 17. Hammett (top) and Taft (bottom) plots of substituted
diazenes. Reaction conditions can be found in Table S4.
4. CONCLUSION
58
however, VTNA of (p-F3COPh)2N2 showed an identical
In conclusion, experiment and theory have shown that Ti-
catalyzed [2+2+1] synthesis of pyrroles from alkynes and
aryl diazenes proceeds through a unique electrocyclic
reductive elimination mechanism. The crux of this
mechanism is that low valent TiII intermediates can be
stabilized both through backdonation into the reaction
products and by solvent effects. In all cases, 2nd alkyne
insertion into an titanacyclobutene is rate determining,
rate equation to the parent azobenzene. Due to the
complexity of this reaction and the fact that NR species
are involved in many steps leading up to the rate
determining step, it is difficult to determine the cause of
this large deviation for strongly EWGs.
To further investigate the ability of the diazene to affect
the reaction rates, we compared 4 different o-substituted
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 20
6)Gilbert, Z. W.; Hue, R. J.; Tonks, I. A. Nature Chemistry
although complex equilibria prior to the rate determining
1
2
3
4
5
6
7
8
2015, 8, 63.
step can significantly alter the reaction rate and
selectivity. Importantly, computational analysis has
allowed us to separate out the stereoelectronic effects that
influence reaction selectivity in each step involving alkyne
reactivity ([2+2] cycloaddition and 1,2-insertion), which
will allow for the new development of ligands and
catalysts capable of controlling regioselectivity of
multisubstituted pyrrole synthesis. Additionally, a key
feature of this reaction is that there is no single reaction
component that is critical in stabilizing intermediates or
transition states throughout the catalytic cycle, indicating
that it should be possible to translate this type of
reactivity beyond simple pyrrole synthesis and into more
broad classes of oxidative catalysis.
7) a.) Müller, T.E.; Hultzsch, K.C.; Yus, M.; Foubelo, F.;
Mizuki, T. Chem. Rev. 2008, 108, 3795-3892 and references
therein. b.) Clough, B. A.; Mellino, S.; Protchenko, A. V.;
Slusarczyk, M.; Stevenson, L. C.; Blake, M. P.; Xie, B.; Clot,
E.; Mountford, P. Inorg. Chem. 2017, 56, 10794.
8)Ryken, S.A.; Schafer, L.L. Acc. Chem. Res. 2015, 48, 2576-
2586.
9)Odom, A.L.; McDaniel, T.;J. Acc. Chem. Res. 2015, 48, 2822-
2833.
10) Guo, J.; Deng, X.; Song, C.; Lu, Y.; Qu, S.; Dang, Y.; Wang,
Z.-X. Chemical Science 2017, 8 (3), 2413-2425.
11) Knizia, G. Journal of Chemical Theory and Computation
2013, 9 (11), 4834-4843;
12) Knizia, G.; Klein Johannes, E. M. N. Angewandte Chemie
International Edition 2015, 54 (18), 5518-5522.
13) a.) Glendening Eric, D.; Landis Clark, R.; Weinhold, F.
Journal of Computational Chemistry 2013, 34 (16), 1429-
1437. b.) Weinhold, F.; Landis, C. R.: Valency and
Bonding: A Natural Bond Orbital Donor-Acceptor
Perspective; Cambridge University Press: Cambridge,
2005.
14) Whitesides, G.M.; Ehmann, W.J. J. Am. Chem. Soc. 1970,
92, 5652-5640
15) Blake, A.J.; Collier, P.E.; Dunn, S.C.; Li, W.-S.; Mountford,
P.; Shishkin, O.V. J. Chem. Soc., Dalton Trans. 1997, 0,
1549-1558.
16) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G.
E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone,
V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato,
M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;
Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota,
K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.;
Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery,
J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.;
Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi,
R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J.
C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J.
M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo,
C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev,
O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G.
A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels,
A. D.; Farkas; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J., Gaussian 09, Revision B.01. Wallingford CT,
2009.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information
Full experimental, kinetic, and computational details are
provided in the supporting information (PDF).
AUTHOR INFORMATION
Corresponding Author
* jgoodpas@umn.edu, *itonks@umn.edu
Author Contributions
‡These authors contributed equally. All authors have given
approval to the final version of the manuscript.
ACKNOWLEDGMENT
Financial support was provided by the University of
Minnesota (start-up funds, Doctoral Dissertation Fellowship
to ZWD-G), the National Institutes of Health (1R35GM119457)
and the Alfred P. Sloan Foundation (IAT is a 2017 Sloan
Fellow). Equipment for the Chemistry Department NMR
facility were supported through a grant from the National
Institutes of Health (S10OD011952) with matching funds from
the University of Minnesota. The authors acknowledge the
Minnesota Supercomputing Institute (MSI) at the University
of Minnesota, and the National Energy Research Scientific
Computing Center (NERSC), a DOE Office of Science User
Facility supported by the Office of Science of the U.S.
Department of Energy under Contract No. DE-AC02-
05CH11231, for providing resources that contributed to the
results reported within this paper.
17) Zhao, Y.; Truhlar, D. G. Theoretical Chemistry Accounts
2008, 120 (1), 215-241.
REFERENCES
1) Hunt. A. Element Recovery and Sustainability, RSC
Publishing: Cambridge, 2013
2) Odom, A.; McDaniel, T.J. Acc. Chem. Res. 2015, 48, 2822
3) Chirik, P.; Morris, R. Acc. Chem. Res. 2015, 48, 2495-2495.
4)Rodriguez-Ruiz, V.; Carlino, R.; Bezzenine-Lafollèe, S.; Gil,
R.; Prim, D.; Schulz, E.; Hannedouche, J. Dalton Trans.
2015, 44, 12029-12059.
18) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. The
Journal of chemical physics 1980, 72 (1), 650-654.
19) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. The Journal
of Physical Chemistry B 2009, 113 (18), 6378-6396.
20)Werner, H. J.; Knowles Peter, J.; Knizia, G.; Manby
Frederick, R.; Schütz, M. Wiley Interdisciplinary Reviews:
Computational Molecular Science 2011, 2 (2), 242-253.
21) Weigend, F.; Ahlrichs, R. Physical Chemistry Chemical
Physics 2005, 7 (18), 3297-3305.
5) Davis-Gilbert, Z.W.; Tonks, I.A. Dalton Trans. 2017, 46,
11522-11528.
ACS Paragon Plus Environment

Page 19 of 20
Journal of the American Chemical Society
22) Webster, A. J.; Mueller, C. M.; Foegen, N. P.; Sit, P. H. L.;
41) Negishi, E.; Takasji, T. Acc. Of Chem. Res. 1994, 24, 124-130
1
2
3
4
5
6
7
8
Speetzen, E. D.; Cunningham, D. W.; D’Acchioli, J. S.
Polyhedron 2016, 114, 128-132.
23) Blackmond, D.J. Angew. Chem. Int. Ed. 2005, 44, 4302-
4320.
42)Epstein, O.L.; Savchenkio, A.I.; Kulinkovich Russian
Chem. Bull. 2000, 49, 378-380.
43) Siebeneicher, H.; Doye, S. J. Prakt. Chem. 2000, 342, 102-
106.
44)Kulinkovich, O.G.; Sviridov, S.V.; Vasilevski, D.A.
Synthesis, 1991, 234.
45)Davis-Gilbert, Z. W.; Yao, L. J.; Tonks, I. A. J. Am. Chem.
Soc. 2016, 138, 14570.
46)Hicks, F.A.; Kablaoui, N.M.; Buchwald S.L. J. Am. Chem.
Soc. 1999, 121, 5881-5898
47)Hicks, F.A.; Buchwald, S.L. J. Am. Chem. Soc. 1999, 121,
7026-7033
48)Marenich, A. V.; Cramer, C. J.; Truhlar, D. G., The Journal
of Physical Chemistry B 2009, 113, 4538-4543.
49)White, D.P.; Anthony, J.C.; Oyefeso, A.O. J. Org. Chem.
1999, 64, 7707-7716
50) Eliel, E.L.;vWilen, S.H.; Mander, L.N. Stereochemistry of
Organic Compounds, Wiley: New York, 1994
51) Hansch, C.; Leo, A.; Taft R.W. Chem. Rev. 1991, 91, 165-195.
52) Chiu, H.-C.; Tonks, I.A. Angew. Chem. Int. Ed. 2018
doi:10.1002/anie.201800595
53) Difranco, S.A.; Masciulis, N.A.; Staples, R.J.; Odom, A.L.
Inorg. Chem., 2012, 51, 1187-1200.
54)Bemowski, R.D.; Singh A.K.; Bajorek, B.J.; Deporre Y.;
Odom, A.L.; Dalton Trans. 2014, 43, 12299-12305
55) Billow, B.S.; McDaniel, T.J.; Odom, A.L. Nat. Chem. 2017,
9, 837-842.
56)Falivene, L.; Credendino, R.; Poater, A.; Petta, A.; Serra,
L.; Oliva, R.; Scarano, V.; Cavallo, L. Organometallics
2016, 35, 2286.
57) Schreck, J.O. J. Chem. Edu. 1971, 48, 103-107.
58)Bergon, M.; Calmon, J.-P. Tet. Let. 1981, 22, 937-940.
24)Blackmond, D.G. J. Am. Chem. Soc. 2015, 137, 10825-10866.
25) Burés, J. Angew. Chem. Int. Ed. 2016, 55, 2028-2031
26)Burés, J. Angew. Chem. Int. Ed. 2016, 55, 16084-16087
27) Rosner, T.; Bars, J.L.; Pfaltz, A.; Blackmond, D.G. J. Am.
Chem. Soc. 2001, 123, 1848-1855
28)Hill, D.E.; Pei, Q.-I.; Zhang, E.; Gage, J.R.; Yu, J.-Q.;
Blackmond, D.G. ACS Catal. 2018, 8, 1528-1531.
29)a.) Bennett, J.L.; Wolczanski, P.T. J. Am. Chem. Soc. 1994,
116, 2179-2180. b.) Nielson, A. J.; Glenny, M. W.; Rickard,
C. E. F. J. Chem. Soc. Dalton Trans. 2001, 232-239.
30) Zuckerman, R.L.; Krska, S.W.; Bergman, R.B. J. Am. Chem.
Soc. 2000, 122, 751-761.
31) Pohlki, F.; Doye, S. Angew. Chem. Int. Ed. 2001, 40, 2305-
2308
32) Johnson, J.S.; Bergman, B.G. J. Am. Chem. Soc. 2001, 123,
2923-2924.
33) Straub, B.F.; Bergman, R.G. Angew. Chem. Int. Ed. 2001,
40, 4632-4635.
34) Esteruelas, M.A.; López, A.M.; Mateo, A.C.; Oñate, E.
Organometallics, 2006, 25, 1448-1460.
35) Streuff, J.; Himmel, D.; Younas, S. L. Dalton Trans. 2018,
Advance Article DOI: 10.1039/C8DT00643A
36)See, X.Y.; Beaumier, E.P.; Davis-Gilbert, Z.W.; Dunn, P.L.;
Larsen, J.A.; Pearce, A.J.; Wheeler, T.A.; Tonks, I.A.
Organometallics, 2017, 36, 1383-1390.
37) Laidler, K. J.; and Tweedale, A. In Advances in Chemical
Physics (eds J. O. Hirschfelder and D. Henderson) 2007.
38) Kramers, H. A.,. Physica 1940, 7, 284-304.
39)Aullón, G.; Alvarez, S., Theoretical Chemistry Accounts
2009, 123 (1), 67-73.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
40)Vinogradov, M.G.; Turova, O.V.; Zlotin, S.G. Org. Biomol.
Chem. 2017, 15, 8245-8269, and references therein.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 20
Table of Contents artwork
1
2
3
4
5
6
7
8
9
Ph
N
R
cat.
[py2TiCl2(NR)]2
R
R
2
+
0.5
PhN=NPh
110 o
C
R
R
R
R
New NꢀC ! ꢀbond
R
R
R
LnTiII
LnTiIV
Ph
R
N lone pair/
N
TiꢀN
π
ꢀbond
R
Ph
N
R
R
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
•
•
•
Variable Time Normalization Analysis
Unsymmetric Alkyne Regioselectivity
Hammett Analysis
•
•
•
Density Functional Theory
Natural Bond Orbital (NBO) Analysis
Intrinsic Bond Orbital (IBO) Analysis
ACS Paragon Plus Environment