Mechanism of the Selective Fe-Catalyzed Arene Carbon-Hydrogen Bond Functionalization

Article abstract of DOI:10.1021/acscatal.7b03935

The complete chemoselective functionalization of aromatic C(sp²)-H bonds of benzene and alkyl benzenes by carbene insertion from ethyl diazoacetate was unknown until the recent discovery of an iron-based catalytic system toward such transformation. A Fe(II) complex bearing the pytacn ligand (pytacn = L1 = 1-(2-pyridylmethyl)-4,7-dimethyl-1,4,7-triazacyclononane) transferred the CHCO₂Et unit exclusively to the C(sp²)-H bond. The cycloheptatriene compound commonly observed through Buchner reaction or, when employing alkyl benzenes, the corresponding derivatives from C(sp³)-H functionalization are not formed. We herein present a combined experimental and computational mechanistic study to explain this exceptional selectivity. Our computational study reveals that the key step is the formation of an enol-like substrate, which is the precursor of the final insertion products. Experimental evidences based on substrate probes and isotopic labeling experiments in favor of this mechanistic interpretation are provided.

Full text of DOI:10.1021/acscatal.7b03935

Subscriber access provided by YORK UNIV
The Mechanism of the Selective Fe-Catalyzed
Arene Carbon-Hydrogen Bond Functionalization
Verònica Postils, Monica Rodriguez, Gerard Sabenya, Ana Conde, M. Mar
Diaz-Requejo, Pedro J. Pérez, Miquel Costas, Miquel Solà, and Josep M. Luis
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b03935 • Publication Date (Web): 20 Mar 2018
Downloaded from http://pubs.acs.org on March 21, 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 12
ACS Catalysis
1
2
3
4
5
6
7
8
9
The Mechanism of the Selective Fe-Catalyzed Arene Carbon-
Hydrogen Bond Functionalization
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
Verònica Postils,†,‡ Mònica Rodríguez,†,‡ Gerard Sabenya,† Ana Conde,§ M. Mar Díaz-Requejo,§,*
Pedro J. Pérez,§,* Miquel Costas,†,* Miquel Solà, †,* and Josep M. Luis†,*
†Institut de Química Computacional i Catàlisi (IQCC) and Departament de Química, Universitat de Girona, Campus Montili-
vi, 17071, Girona, Spain
§Laboratorio de Catálisis Homogénea, Unidad Asociada al CSIC, CIQSO-Centro de Investigación en Química Sostenible and
Departamento de Química, Universidad de Huelva, 21007 Huelva, Spain
ABSTRACT: The complete chemoselective functionalization of aromatic C(sp2)-H bonds of benzene and alkyl-benzenes by car-
bene insertion from ethyl diazoacetate was unknown until the recent discovery of an iron-based catalytic system toward such trans-
formation. A Fe(II) complex bearing the pytacn ligand (pytacn=L1=1-(2-pyridylmethyl)-4,7-dimethyl-1,4,7-triazacyclononane)
transferred the CHCO2Et unit exclusively to the C(sp2)-H bond. The cycloheptatriene compound commonly observed through
Buchner reaction or, when employing alkyl-benzenes, the corresponding derivatives from C(sp3)-H functionalization are not
formed. We herein present a combined experimental and computational mechanistic study to explain this exceptional selectivity.
Our computational study reveals that the key step is the formation of an enol-like substrate, which is the precursor of the final inser-
tion products. Experimental evidences based on substrate probes and isotopic labelling experiments in favor of this mechanistic
interpretation are provided.
KEYWORDS carbene transfer iron catalysis C-H activation C(sp2)-H functionalization DFT calculations
so-called Buchner reaction, known in a thermal manner since
the XIX century,2 and in the rhodium-catalyzed version after
the work by Noels and co-workers.3 The second transfor-
mation is that of the neat insertion4 of the carbene unit in the
C(sp2)-H bond of the arene, a less frequent transformation. To
date the functionalization of benzene and other non-
substituted arenes by this methodology based on metal-
carbene intermediates, generated from diazo reagents, has
been achieved using several metals (Scheme 2) such as rhodi-
um,5 or coinage metals.6,7 In all these cases, a mixture of
products is obtained due to the existence of aforementioned
competing reactions (Scheme 1). Until recently, the best se-
lectivity for insertion as opposed to addition products is ca 3:1
(Scheme 2a), and was obtained with a gold-based catalyst.7 A
chemoselective insertion of the carbene group in aromatic
C(sp2)-H bonds can be achieved with the use of electron-rich
benzene derivatives such as phenol or anisole (Scheme 2b).
The addition of these π-electron-donating groups (-OH,
-OCH3) to benzene enhances the nucleophilic reactivity of the
INTRODUCTION
The metal-catalyzed carbene transfer from a diazo reagent to
C-H bonds constitutes a strategy that is gaining interest in the
last decade.1 Both C(sp2)-H and C(sp3)-H bonds can be modi-
fied by this methodology, albeit when working with benzene
derivatives the issue of selectivity emerges as a serious draw-
back. This is the consequence of a series of competing reac-
tions that may occur, as shown in Scheme 1. The addition of
the carbene unit to a C=C bond of the arene leads to a norca-
radiene intermediate that spontaneously converts into the
cycloheptatriene derivative, the addition product. This is the
Scheme 1. Possible products in the metal catalyzed
(alkyl)-benzene functionalization by carbene insertion
8
aromatic carbons and favors insertion products. The use of
electron-deficient phosphite ligands increases the electrophilic
character of the carbene moiety in gold catalysts, enabling
selective insertion in electron-rich arenes. Finally, the exist-
ence of alkyl-substituents9 or other X-H groups (X = O, H)10
in the substrate may lead to their functionalization, also upon
insertion of the carbene group. These substrate functionaliza-
tion reactions also compete with the non-productive metal-
catalyzed coupling of two carbene units.11
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 12
1
2
Scheme 2. Functionalization of arenes by metal-carbene catalysts.
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
a) Precedents in the catalytic functionalization of benzene by metal-carbene transfer. Non-chemoselective processes. b) Precedents in
the functionalization of activated benzenes (phenol) by metal-carbene transfer. Chemoselective processes. c) Reaction of this study.
First example of a chemoselective functionalization of benzene.
under N2, and were initiated by the addition of 20 equiv of
EDA to a solution of 5 mol% of the catalyst and NaBAr4F (2-8
equiv). Reactions provided insertion (I) and addition (A)
products, and the yields and relative amount of both products
were dependent on the catalyst. The typology of the catalysts
studied in our previous work was mainly focused in tetraden-
tate aminopyridine ligand (Scheme 3), which have proven to
form iron complexes particularly active in oxygen atom trans-
fer reactions (L1 and L7 constitute prototypical examples).13
In the current work, this collection has been extended with the
use of iron complexes bearing sterically encumbered triden-
tate (L10) and tetradentate (L9) ligands based in pyrazole
rings, and that have been found particularly active in other
carbene transfer reactions with first row transition metals.14
This analysis reinforces [FeX2(L1)] (where X = Cl or OTf) as
a particularly efficient (entries 1 and 2, up to 86% of EDA
was converted into products) and chemoselective catalysts
towards the insertion product (>99%). Modifications in the
electron donating properties of the pyridine (entries 3-6), and
in the denticity of the ligand (entries 7 and 9) led to lower
product yields. Particularly noticeable was the observation
that complex [Fe(OTf)(L8)](OTf), bearing a pentadentate
ligand, was inactive (entry 9), strongly suggesting that the
presence of two labile sites at the first coordination sphere of
the iron center is necessary for activity. Complexes based in
pyrazolyl donors, [Fe(OTf)2(L9)] and [Fe(OTf)(L10)] are
active (entries 10 and 11) but provide roughly 1:1 mixtures of
insertion and addition products.
Recently, we reported the first example for a chemoselective
functionalization of non-activated aryl C(sp2)-H bonds with
the use of the commercially available ethyl diazoacetate
(EDA) and iron- or manganese-based catalysts.12 FeII and MnII
complexes bearing the tetradentate pytacn ligand (pytacn = L1
= 1-(2-pyridylmethyl)-4,7-dimethyl-1,4,7-triazacyclononane)
lead to >99% formation of insertion products and no for-
mation of cycloheptatriene derivatives (Scheme 2c). Fumarate
and maleate products, resulting from coupling of two carbene
units, were also practically absent. This work also expanded to
first-row transition metal catalysts the direct functionalization
of non-activated aryl C(sp2)-H bonds by metal-carbene trans-
fer from ethyl diazoacetate. Interestingly, the presence of
substituents in a series of alkyl-aromatic substrates did not
affect the selectivity of the reaction, since the alkylic C-H
bonds remained unreacted, only the aryl C-H bonds being
functionalized. In view of the novelty of this transformation in
terms of the exceptional selectivity, we have performed a
combined experimental and theoretical study that has led to
the understanding of the mechanistic details of the functional-
ization of C-H bond of benzene, which may serve in the de-
sign of novel generations of catalysts with improved efficien-
cy and chemoselectivity.
RESULTS AND DISCUSSION
Experimental analyses
On the basis of the previous observations, [Fe(OTf)2(L1)]
was taken as the catalyst of choice to test other carbene
sources, that led to limited success (entries 12 and 13). Unlike
EDA, which provides insertion products in good yield and
selectivity, the use of phenyldiazoacetate of its 4-bromo de-
Effect of the catalyst architecture in the catalytic activi-
ty. Initially we studied the potential ability of iron complexes
to catalyze the reaction of ethyl diazoacetate (EDA) and ben-
zene (Scheme 2c and Table 1). Reactions were performed at
80ºC during 12h in a 1:1 mixture of benzene and CH2Cl2
2
ACS Paragon Plus Environment

Page 3 of 12
ACS Catalysis
rivative resulted instead in a slight preference towards the
expansion product in moderate (41-51%) yields.
formed when this salt reacts with [Fe(OTf)2(L1)]. Diiron
F
complex [Fe2(µ2-OTf)2(L1)2](BAr4 )2, was prepared by reac-
1
2
F
tion of [Fe(OTf)2(L1)] with 1-4 equiv. of NaBAr4 , and it was
crystallographically characterized. An ORTEP diagram of the
diiron complex is shown in Figure 1, and compared with that
of [Fe(OTf)2(L1)].15 Iron centers in both complexes are coor-
dinatively saturated. They have a distorted octahedral coordi-
nation geometry, with four sites occupied by the L1 ligand
and the two remaining sites fulfilled by oxygen atoms of the
triflate ligands, which are terminally coordinated in the mono-
nuclear complex and bridging in the dinuclear case.
3
4
5
6
7
8
9
Table 1. Catalytic behavior of different
[FeX2(L)] complexes in the reaction of benzene
and ethyl diazoacetate.a
Entry
L
X
Yieldb
%I:%Ac,d
Ref
1
L1
L1
L2
L3
L4
L5
L6
L7
L8
L9
L10e
L1
L1
Cl
86
83
77
47
85
18
67
70
n.r
57
55
41
51
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
>99:1
…
12
2
OTf
OTf
OTf
OTf
OTf
OTf
OTf
OTf
Cl
12
12
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
When [Fe(OTf)2(L1)] complex was tested under standard
F
3
reaction conditions, but in the absence of additional NaBAr4 ,
it was catalytically inactive. However, when the same reaction
4
12
F
was conducted in the presence of 4 equiv. of NaBAr4 , the
insertion product was obtained as the only product in 80%
yield. These observations strongly suggest that detachment of
the first triflate anion from [Fe(OTf)2(L1)] does not lead to an
active catalyst, and instead removal of the two anionic lig-
ands, generating two vacant sites at the iron coordination
sphere, are most likely needed for EDA activation. Unfortu-
nately, iron species resulting from elimination of the second
triflate ligand could not be isolated, and the paramagnetic
nature of the iron species makes 1H-NMR ineffective to probe
its formation in solution. However, prominent ions at m/z =
152.1 with isotopic patterns characteristic of [FeL1)]2+ could
be observed in the ESI-MS spectra of [Fe(OTf)2(L1)], sug-
gesting that tetracoordinate dicationic species are viable in-
termediates.
5
12
6
12
7
12
8
12
9
12
10
11
12
13
42:58
53:47
45:55f
40:60g
This work
This work
This work
This work
Cl
OTf
OTf
aReactions carried out at 80°C with 0.005 mmol of catalyst, 8
equiv of NaBAr4F and 20 equiv of ethyl diazoacetate in 1.5 mL of
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
b
benzene + 1.5 mL of CH2Cl2. Reaction time: 12 h. Initial EDA
converted into insertion (I) + addition (A) products. Determined
by GC; some ethyl glycolate from adventitious water was detected
c
d
as byproduct. Determined by NMR. The acid-catalyzed conver-
sion of A into I has been discarded on the basis of blank experi-
f
ment: see Supporting Information. eFeX[L10] complex Phenyldi-
g
azoacetate was used as carbene source instead of EDA. 4-Br-
phenyldiazoacetate was used as carbene source instead of EDA.
Figure 1. ORTEP diagram of the single crystal X-ray struc-
Scheme 3. Complexes employed as catalyst precursors in
the screening shown in Table 1.
F
ture for [Fe(OTf)2(L1)] (left) and [Fe2(µ2-OTf)2(L1)2](BAr4 )2
(right).15 50% ellipsoid probability, H atoms have been omit-
ted for clarity. [Fe(OTf)2(L1)] selected distances (Å): Fe1-N1:
2.231, Fe1-N2: 2.251, Fe1-N3: 2.205, Fe1-N4: 2.165, Fe1-O1:
F
2.055, Fe1-O2: 2.165. [Fe2(µ-OTf)2(L1)2](BAr4 )2 selected
distances (Å): Fe1-N1: 2.200, Fe1-N2: 2.210, Fe1-N3: 2.187,
Fe1-N4: 2.142, Fe1-O1: 2.060, Fe1-O2: 2.189.
The results collected in this section reveal that
[Fe(OTf)2(L1)] is a particularly efficient and chemoselective
catalyst for the formal carbene insertion from EDA into the C-
H bonds of benzene, reinforcing the findings of the prelimi-
nary communication.12 It is worth noting that the scope of this
catalytic system with a series of substituted benzenes showed
the complete selectivity toward such bonds in the presence of
alkyl C-H sites of the substituents. The presence of two labile
coordination sites appears to be necessary for efficient cataly-
sis.
F
Mechanistic considerations from product analysis. Our
previous contribution included12 a Hammett analysis of the
carbene transfer reaction, determined by means of competitive
Considering that 2-8 equivalents of NaBAr4 are necessary
for efficient activity, we investigated the iron species that are
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 12
functionalization of pairs of monosubstituted arenes, which as the major product, presumably because: i) C5 and C6 are
1
2
3
4
5
6
7
8
provided ρ = -2.75 ± 0.37, indicative of an electrophilic func-
tionalization of the arene. Furthermore, an inverse intermolec-
ular KIE = 0.95 was determined, consistent with a change in
the hybridization from sp2 to sp3 in the arene functionalization
step. These parameters strongly suggest that the functionaliza-
tion of the arene occurs via an electrophilic aromatic substitu-
tion, a proposal that is in good concordance with the o:m:p
ratios observed in the functionalization of monosubstituted
benzenes. We have now collected additional pieces of infor-
mation en route to mechanism elucidation.
sterically less congested than C3; and ii) C5 is preferentially
activated towards an electrophilic attack because it is in para
position with respect to the methyl group at C2. We note that
4b, 4c, 4d and 4e may be also at least partially originating
from reactions where methyl migration takes place (see Figure
2).
Scheme 5. Carbene insertion into radical-clock substrate
probes.
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
Scheme 4. Alkyl migrations in the functionalization of
1,3,5-trialkylsubstituted arenes.
Alkyl group migrations. Functionalization of mesitylene (2)
and 1,3,5-triethylbenzene (3) under standard reaction condi-
tions (Scheme 4) provides the ethyl esters 2a and 3a as major
products resulting from a formal insertion at arene C-H bond.
However, inspection of the reaction crude NMR spectra reveal
the formation of a minor product 2b and 3b, respectively
(~11%) resulting from alkyl migration.
Radical clock substrate probes. An experiment of this type
was provided in our previous report, where we described that
the functionalization of cyclopropylbenzene (5) with
[Fe(OTf)2(L1)] proceeds without ring opening (Scheme 5a).
This experiment suggested that radical intermediates were not
formed, or alternatively they must have very short lifetimes,
incompatible with radical rearrangements. To further substan-
tiate this point, an ultrafast radical probe (6, Scheme 5b) was
subjected to standard reaction conditions, employing the same
catalysts. The reaction produced not-rearranged cyclopropyl
product 6a, and rearranged product 6b in a 68:32 ratio, with a
combined 39% yield. This observation may initially suggest
that the reaction proceeds via short living carbocationic or
radical intermediates. Most remarkably, an almost identical
ratio (65:35) and yield (40%) was obtained when
[Zn(OTf)2(L1)] was used as catalyst (see Supporting Infor-
mation), suggesting that carbocationic and not radical rear-
rangements are involved.16
These results demonstrate that the reaction mechanism for
carbene insertion must accommodate reaction intermediates
susceptible to engage alkyl arene migrations and carbocationic
rearrangements. Indeed, we notice that migrations may be also
indicative of carbocationic intermediates.
Figure 2. 1H-NMR spectrum (CDCl3, 400MHz, 300k) of the
EDA insertion products obtained in the functionalization of 4
with [Fe(OTf)2(L1)]. The origin of the different products is indi-
cated in each case.
Isotopic labelling analyses. Isotopic labelling analyses has
revealed valuable mechanistic details in order to validate the
computational analysis that follows. Under standard reaction
conditions, formal carbene transfer to 1,3,5-D3-benzene D3-1
yields D3-1a, resulting from formal carbene insertion into an
arene C-H bond, and D3-1b, where insertion has occurred at
the C-D bond (Scheme 6a). The ratio of insertion vs. expan-
sion is higher than 99%, indicating that the deuteration of
benzene does not change the chemoselectivity of the catalysis.
Interestingly, partially D atom ends up forming the benzylic
C-D bond in the D3-1b product. A 8% of the D is incorporated
Alkyl migrations are best evidenced by performing the func-
tionalization of 1,2,4-trimethylbenzene (4) under standard
reaction conditions (Figure 2). The reaction produced three
major (4a, 4b and 4c) and two minor (4d and 4e) products.
Major products originate from formal carbene insertion in C-
H bonds at C5, C3 and C6, respectively, while minor products
are indicative of insertions accompanied by alkyl migrations.
4d can be explained based on an initial attack at C1, followed
by methyl migration to C6, while 4e indicates initial attack on
C4 followed by methyl migration to C3. Of these, 4a appears
4
ACS Paragon Plus Environment

Page 5 of 12
ACS Catalysis
in the product, with the remaining corresponding to the fully
interacts with the iron (Figure 3). The carbonyl O-bound
adduct A is the most stable form, while the carbon-bound
adduct A’ is 14.2 kcal/mol higher in energy. Then, the most
stable adduct is given by the interaction between the iron and
the most nucleophilic atom of EDA. The nitrogen-bound
adduct has not been characterized as a stable minima, as op-
posite to other metal catalysts interacting with EDA.18 For
[Fe(L1)]2+ and all EDA-bound adducts, the most stable spin
state is the quintet (see Figure S1 for more details).
1
2
3
4
5
6
7
8
hydrogenated benzylic position, suggesting the possibility of
the incorporation of protons from an external source. The ratio
between the two mentioned products is 49.3:50.7 (extracted
from relative integration of the aromatic signals), which trans-
lates into an intramolecular kinetic isotope effect, KIE = 0.97,
in agreement with the intermolecular KIE previously deter-
mined from the competitive functionalization of benzene and
D6-benzene. Noticeable, products resulting from deuterium
migration in the aromatic ring, analogous to those observed
from alkyl groups, are not observed.
Only the A’ adduct is able to evolve, via N2 elimination, to-
wards the formation of the Fe-carbene moiety. The relative
instability of A’ with respect to A leads to a global energy
barrier of 29.5 kcal/mol for the overall reaction (Figure 3).
Although an energy barrier of 29.5 kcal/mol is not affordable
at room temperature, it can be overcome from ca. 80°C, which
is the experimental working temperature. High-energy barriers
for carbene formation were also found for the formation of
iron and cobalt porphyrin carbenes.19 This, along with the
formation of the iron-carbene in the rate-determining step of
the overall reaction of benzene functionalization, prevents the
accumulation and experimental characterization of the puta-
tive iron-carbene or any other intermediate. Thus, the compu-
tational study becomes necessary for a proper mechanistic
description.
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
Finally, when the functionalization of benzene was conduct-
ed in the presence of D2O (0.25 equiv.), 42% of the corre-
sponding benzyl ester product 1a contains a benzylic C-D
group (Scheme 6b). Control experiments showed that the
benzylic C-H bonds in 1a does not exchange the hydrogen
atoms with D2O. Consequently, these experiments indicate
that the carbene transfer reaction involves species that could
exchange an H atom with D2O. The involvement of adventi-
tious water in these transformations with gold-catalysts has
been recently proposed.6
Scheme 6. Isotopic labelling analysis.
Once the transition state (TS) for the S=2 carbene formation
(TSA’-C@N2) is surmounted, a spin crossing takes place and
triplet becomes the most stable spin state for the adduct be-
tween the Fe-carbene and N2 (C@N2). Finally, the loss of the
N2 moiety is favorable and the global carbene formation pro-
cess (A to C) is endoergic by 4.9 kcal/mol. A quasi-equivalent
C@N2 isomer is obtained if EDA is coordinated to the labile
position perpendicular to the pyridyl ring instead of the paral-
lel one (see Figure S2).
The above reactivity implies that the mechanism of carbene
insertion must accommodate a path that enables the transfer of
an arene hydrogen atom to the benzylic position, and partial
exchange of a benzylic C-H bond with external water mole-
cules.
25.3
TSA'-C@N2
Computational analyses
Generation of the carbene moiety. It is widely accepted
that the most feasible reaction of metal catalysts with EDA
proceeds via N2 loss to form a metal-carbene species.17 We
have studied the mechanism of this reaction for
[Fe(OTf)2(L1)], taking into account the singlet, triplet, and
quintet spin states. "Since the experimental data shows that
the diiron complex [Fe2(µ2-OTf)2(L1)2]2+ does not participate
in the activation of EDA, it was not considered in the compu-
tational analyses" All structures were optimized and character-
ized at the UB3LYP-D3BJ/6-31G(d)/SMD(CH2Cl2) level of
theory. Then, the electronic energies were improved perform-
ing single point calculations at the UB3LYP-D3BJ/cc-
pVTZ/SMD(50%Bz,50%CH2Cl2) level (more details about
computational details can be found in the SI).
13.6
10.5
10.0
S=0
A'
5.4
C-coordinated-EDA
5.9
N2
S=1
C@N2
0.7
C
0.0
ΔG (carbene formation)=
29.5 kcal/mol
B
[(pytacn)FeII
+2
]
S=2
+ EDA
-4.2
A
O-coordinated-EDA
Figure 3. Gibbs energy profile, ΔG, (kcal·mol-1) of the met-
al-carbene bond formation. Blue, black and green profiles
correspond to quintet, triplet and singlet multiplicities, respec-
tively. Hydrogen atoms of L1 ligand have been omitted for
clarity (C: grey, N: blue, O: red, Fe: yellow, H: white).
Based on the previously described experimental data, which
suggests that the two triflate ligands are removed in the cata-
lytically active species, two possible [(L)Fe-(EDA)]2+ adducts
have been characterized depending on which atom of EDA
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 12
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 4. Gibbs energy profile, ΔG (kcal·mol-1) of the C-to-C’ conformational change at the triplet spin state. For C and C’ interme-
diates, the Fe-C bond distance is given in Å. Hydrogen atoms of the L1 ligand have been omitted for clarity (C: grey, N: blue, O: red,
Fe: yellow, H: white).
The metal center in the iron-carbene species (C) displays a
distorted octahedral coordination geometry, with one free-
coordination site and five occupied-coordination sites taken
by the tetradentate ligand and the carbene. The Fe-Ccarbene
distance in C is 1.85 Å (Figure 4). Iron-carbene bond lengths
are sensitive to the electronic nature and chemical environ-
ments of the carbene and iron ligands.20 The DFT analyses of
the spin densities, frontier molecular orbitals (MOs), natural
bond orbitals (NBO),21 and the effective oxidation states (ef-
fOS)22 suggest that the electronic nature of C Fe-carbene is
described as a Fe(III) and a formal radical alkyl substituent
(see Supporting Information for details). An equivalent elec-
tronic structure was reported by Shaik et al. for an iron por-
phyrin carbene complex.17b
thermore, NCI analyses evidence that the C’ conformer pre-
sents a less crowded environment than C for Ccarb (see section
3.2 in SI for the detailed NCI analysis). Therefore, on the
contrary to C, C’ has enough vacant space to allow: i) the
approach of benzene in front of the free-coordination site of
the iron; and ii) the movement of the carbene hydrogen re-
quired for the interaction between benzene and Ccarb (see
Figure S7). Thus, the comparison between C and C’ also
shows the essential role of the free-coordination site of the Fe-
carbene in the functionalization of benzene, which agrees with
the experimental need to add more than 2 equivalents of
F
F
NaBAr4 for the reaction to proceed. The excess of NaBAr4
ensures the formation of the initial dicationic species
[Fe(L1)]2+ with two free coordination sites (complex B, Fig-
ure3), required for the formation of the metal-carbene.
Iron-carbene C is not the Fe-carbene conformer that pre-
sents the lowest barrier to react with benzene. Among all the
conformers analyzed, the C’ conformer was found to be the
most suitable for the approach of benzene to the Ccarb and to
facilitate further benzene functionalization reactions. The
main difference between C and C’ is the position of the car-
bene tail (see Figure 4). While in the C conformer the carbon-
yl group is pointing towards the free-coordination site of the
catalyst, in the C’ conformer, the tail of the carbene is bent
towards the equatorial-NCH3 group of the ligand. Although in
the C conformer the carbonyl oxygen of the carbene tail is
pointing toward the free-coordination site of iron, on the basis
of non-covalent interaction (NCI)23 and NBO analysis, a
bonding interaction between the carbonyl oxygen of the car-
bene tail and the iron atom is discarded (see Figure S6). Fur-
The detailed pathway from C to C’ has been computational-
ly studied and it is shown in Figure 4. The evolution contains
several low-energy transition states connecting intermediates
of similar energy that match up to three consecutive σ-bond
rotations24 of: i) the bond of Ccarb with the carbonyl carbon
(TSC-Ca), ii) the bond of Ccarb with the iron (TSCa-Cb), and iii)
again the bond of Ccarb with the carbonyl carbon (TSCb-C’).
TSCa-Cb has the highest energy barrier of the conformational
evolution, which is only 3.4 kcal/mol. The Fe-Ccarb bond rota-
tion and its low energy barrier are consistent with the exist-
ence of a very weak Fe-Ccarb π-interaction. Apart from the key
differences pointed above, conformers C and C’ have similar
geometric and electronic structures (see SI for details).
Scheme 7. Plausible pathways for the formation of insertion and addition products
6
ACS Paragon Plus Environment

Page 7 of 12
ACS Catalysis
ligand have been omitted for clarity (C: grey, N: blue, O: red, Fe:
yellow, H: white).
1
Mechanistic insights in benzene functionalization I. Con-
certed insertion versus stepwise mechanisms. Our study of
the first step of the reaction between the iron-carbene and
benzene has taken into account two different possible path-
ways: an stepwise mechanism and a concerted one (see
Scheme 7 and Figure 5). In the former, the attack of the aro-
matic benzene C(sp2) to Ccarb (TSI_II) leads to an intermediate
with a Ccarb-Cbz bond (II). The intermediate II can evolve
through two different pathways to both possible final prod-
ucts, i.e. the insertion and expansion products (see next sec-
tion). In the concerted mechanism, the attack of the aromatic
benzene C(sp2) to Ccarb (TSConc) directly gives insertion prod-
ucts. Both pathways have been described in the literature on
mechanisms of C-H bond insertion by metal-carbenes.25 The
singlet, triplet, and quintuplet spin states were considered in
the calculation of the Gibbs energy profile. Besides the study
of the reaction of C’ with benzene that yields the adduct I, it
has also been considered the reaction of benzene with a biden-
tate carbene isomer (Iκ2), which is 1.2 kcal/mol less stable
than I. The Iκ2 isomer with a singlet ground state can be un-
derstood as an evolution of carbene C, where the carbonyl
oxygen is bonded to iron through its sixth-coordination site. A
global picture of the results is represented in Figure 5.
2
3
4
5
6
7
8
9
For the concerted mechanism, a high-energy barrier of 25.8
kcal/mol (TSConc) at the triplet spin ground state is found.
Thus, the concerted mechanism is also discarded. In agree-
ment with our results, the fact that metal-carbenes do not
perform a concerted Csubstrate-H bond insertion for aromatic
substrates contrary to alkane substrates has been also recently
reported for gold-carbenes.8a The same conclusion can also be
obtained comparing the mechanistic literature of the reactions
of Au, Cu, Ag or Rh carbenes with alkanes26 and with aro-
matic compounds.6a,25b
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
DFT analysis of the electronic structure of intermediate II
suggests a cyclohexadienyl radical character intermediate (see
Table S7). However, several pieces of experimental evidences
point towards this reaction proceeding via an electrophilic
attack of the carbene on the aromatic ring, generating a
Wheland type intermediate. On one hand, o:m:p ratios and
Hammett parameters are consistent with this scenario. On the
other hand, the inverted KIE effect is in good agreement with
a sp2 to sp3 change in hybridization in the arene functionaliza-
tion step. Furthermore, the cationic nature of the intermediate
is well supported by the observation of alkyl migration prod-
ucts, and by the carbocationic rearrangement product obtained
in the functionalization of the ultrafast radical probe 6. The
degree of reorganization of this substrate is identical to that
measured with the red-ox inactive [Zn(OTf)2(L1)] catalyst,
providing strong argumentation against a radical electronic
structure of II. The radical electronic structure of 3II predicted
by DFT calculations could be due to the particular strong spin
The most favorable mechanism for the initial step of the
functionalization of benzene is the Ccarb-Cbz bond formation
from monodentate C’ to yield II through TSI_II, which is, at
least, 8.5 kcal/mol lower in energy than any other studied
pathway. This step has a global energy barrier of only 17.3
kcal/mol whereas the same reaction mechanism for the biden-
tate carbene has an energy barrier of 25.9 kcal/mol (TSκ2I_II).
Therefore, we discard any role of isomer Iκ2 in the functionali-
zation of benzene (see Figures S10, S12, and S13 in the SI for
geometry and electronic details of TSI_II and TSκ2I_II).
3
contamination of the wavefunction of intermediate II (i.e.
<S2> =3.1). Nevertheless, it is important to remark here that
we have used the Yamaguchi correction27 to remove the spin-
contamination error of the computed Gibbs energies used to
determine the most favorable mechanism (see Computational
Details and Section 6 in the SI. Several conformers for all the
reaction pathways and S=0, S=1, and S=2 spin states were
examined. Figure 5 just shows the most important results and
the complete data is shown in Figure S11 and Tables S4, S5
and S6. Emphasis had been placed on a proper description of
the energy barrier and spin splitting of the Ccarb-Cbz bond
formation (3TSI_II and 5TSI_II). Thus, although the precise
electronic structure of intermediate 3II is not exactly described
by our methodology, additional calculations (vide infra) indi-
cate that the Gibbs barrier of the transformation of I to II
through 3TSI_II is properly estimated.
25.9
2
TSκ
25.8
I_II
TSConc
17.3
TSI_II
13.4
II
5
It has to be mentioned here that TSI_II is found to be 7.8
3.9
S=0
3
2.7
I
2
Iκ
kcal/mol lower in energy than TSI_II. However, we have not
0.7
C
5
0.0
S=1
taken into account TSI_II in our mechanism for the reasons
C'
that follow. It is known that B3LYP is not a good functional
to describe the correlation energy changes of reactions where
the number of unpaired electrons of the metal-carbon bond
changes.28 Furthermore, B3LYP tends to overstabilize the
high-spin states.29 A study of homolytic Fe-, Co-, and Ni-C
bond dissociation energies in tetra-pyrrole-like systems shows
that the use of pure DFT methods with Becke’s exchange
functionals (BP86, BLYP or BPW91) are a better alternative
than B3LYP to describe the metal-C bond breaking.28a Thus,
we have optimized I and TSI_II for the triplet and quintet spin
+
Figure 5. Comparative Gibbs energy profile, ΔG, (kcal·mol-1)
of the first steps of the studied reactions between benzene and the
iron-carbene. Black and green profiles correspond to triplet and
singlet multiplicities, respectively. Hydrogen atoms of the L1
7
ACS Paragon Plus Environment

ACS Catalysis
states using B3LYP, B3LYP*, OPBE, M06-L, BP86, BLYP,
Page 8 of 12
1
2
3
4
5
6
7
8
and BPW91 functionals and 6-31G(d), 6-311+G(d,p) and
Def2-SVP basis sets (see Table S5 and S6 in the SI for de-
tails). The results obtained with the literature-recommended
Two different pathways connect II with insertion products:
i) the classical direct hydrogen migration between the two
carbons that form the Ccarb-Cbz bond (i.e. from Cbz to Ccarb), the
called [1,2]-H shift reaction (TSII_V); and ii) a hydrogen ab-
straction reaction of the same hydrogen by the carbonyl oxy-
gen of the native EDA (TSII_III). The latter mechanism, named
carbonyl assisted insertion, generates an enol intermediate
that, to finally yield the insertion product, should subsequently
evolve to the ester form. On the other hand, we only found
one possible reaction that connects II with addition products.
28
functionals for this step of the mechanism (i.e. BP86, BLYP
and BPW91) are similar among them and indicate that the
triplet state is the spin ground state for both I and TSI_II. Be-
sides, the three recommended functionals lead to similar
3TSI_II energy barriers. Thus, according to these results, we do
3
5
not consider the spin-crossing from TSI_II to TSI_II. Howev-
er, we keep B3LYP as the reference functional because hybrid
methods like B3LYP are more reliable functionals to study the
other steps of the mechanism that involve formation and
breaking of bonds between H, C, and O atoms.30 For con-
sistency, we have chosen the same functional, B3LYP, to
present the results for all the steps of the mechanism.
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 addition pathway implies the formation of a second Ccarb
-
Cbz bond with an adjacent carbon of the phenyl (TSII_IV). This
TS does not directly give the addition product, but a norca-
radiene, which can evolve to the final cycloheptatriene addi-
tion product (VI) through a thermally allowed electrocyclic
disrotatory ring opening.
As it is shown in Figure 6, among the three different path-
ways that start at II, the formation of the enol is the lowest
energy barrier process (TSII_III). The enol formation from II is
strongly exergonic. Then, the next and last transition state of
this pathway, the enol-to-ester transformation barrier, takes
place at far lower Gibbs energy. Thus, the most favorable
mechanism from II is the formation of the enol intermediate
that leads to the final insertion of -C(H)CO2Et into benzene’s
C-H bond. The abstraction of the hydrogen bonded to the
phenylic carbon that forms the Ccarb-Cbz bond by one carbonyl
oxygen has also been reported as the most favorable mecha-
nism for the C-H functionalization of arenes by Au and Ag
carbene catalysts,6,Error! Bookmark not defined. and in the functionali-
zation of N-H bonds by Fe porphyrin carbene catalyst.19b
Mechanistic Insights in the benzene functionalization II.
Insertion versus Expansion mechanisms. In this section the
mechanism that determines the complete selective insertion of
-C(H)CO2Et into benzene’s C-H bonds with no formation of
cycloheptatriene byproducts is addressed. Figure 6 presents all
the possible different mechanisms we found that connect
intermediate II with insertion or addition products.
22.8
TSII_V
(TS1,2H-shift
)
16.3
TSII_IV
13.4
II
14.6
TSII_III
The enol formation process has a small energy barrier of 1.2
kcal/mol. Although the enol product formed still remains
close to iron (the Fe···C distance is 2.28 Å), there is not a
formal bond between them and iron is described as an
Fe2+(d6). The enol intermediate structure III that involves a
high spin d6 iron with four unpaired electrons, 5III, is energet-
5.3
-7.1
TSIV_VI
-9.7
-14.7
VI
IV
-18.6
-20.7
S=2
-30.7
TSIII_V(2w)
III
3
ically more favorable than III by 12.1 kcal/mol. Thus, a spin
crossing takes places at this stage and the global process be-
comes highly exoergic (44.1 kcal/mol) because the benzyl
group recovers the aromatic character.
-60.2
V
The formation of the norcaradiene, the addition-product pre-
cursor, through TSII_IV and the direct formation of the inser-
tion product through the classical hydrogen migration from
the phenylic carbon to the benzylic carbon, TSII_V, involve
energy barriers of 2.9 and 9.4 kcal/mol, respectively. Thus,
+2
+2
H
H
H
H
H
H
H
H
H
H
H
H
2·H2O
H
H
H
(L1)FeII
H
H
C
H
OH
(L1)Fe
C
C
(L1)Fe
C
CO2Et
O
C
H
OEt
H
OEt
III
II
both TSII_IV and TSII_V have higher energies than TSII_III
,
H
H
+2
H
H
H
which implies that these other two reactions will hardly take
place. This result nicely agrees and explains the experimental
data, which show that our Fe catalyst presents a clear selectiv-
ity for the generation of insertion products. It is worth noticing
that, in electronic energy terms, the enol formation barrier is
higher than the norcaradiene formation barrier. Then, entropy
has a key effect favoring TSII_III over TSII_IV due to the loose
five-membered ring structure of TSII_III (see Table S8). When
EDA is replaced by 4-Br-phenyldiazoacetate (entry 13 of
Table 1), DFT barriers show that the chemoselectivity for the
formation of insertion products decreases (see Table S11), in
nice agreement with the experimental data and the reduction
of the carbonyl group basicity of II.
H
O
H
H
H
H
H
H
H
H
O
H
C
(L1)FeII
(L1)FeII
H
C
H
H
O
O
C
C
H2
HHO
O
OEt
OEt
V
Figure 6. Comparative Gibbs energy profile, ΔG, (kcal·mol-1)
of the formation of addition (VI) and insertion (V) products from
II. Black and blue profiles correspond to triplet and quintet mul-
tiplicities, respectively. Hydrogen atoms of the L1 ligand have
been omitted for clarity (C: grey, N: blue, O: red, Fe: yellow, H:
white). The inset below the energy profile details the reaction
mechanism to obtain insertion products through the formation of
the enol intermediate.
8
ACS Paragon Plus Environment

Page 9 of 12
ACS Catalysis
Both the norcaradiene-formation transition state and the
ASSOCIATED CONTENT
Supporting Information
1
2
3
4
[1,2]-H shift transition state imply the rearrangement reaction
and the dissociation of the products formed. In TSII_IV, besides
the substrate formation, norcaradiene decoordinates from the
Experimental procedures and DFT calculations are provided as a
single pdf file. Cartesian coordinates of all computed structures
are given in a separate text file in xyz format. The Supporting
Information is available free of charge on the ACS Publications
website.
catalyst (Fe···Ccarb distance of IV is 3.30Å); and, in TSII_V
,
along with the substrate formation and decoordination
(Fe···Ccarb distance of V is 3.62Å), the formation of a new
adduct between Fe and the oxygen atom of the carbonyl group
of insertion products occurs (Fe···Ocarb distance of V is
1.97Å). Despite the Fe···Ocarb interaction in IV and V prod-
ucts, iron recovers its Fe2+(d6) electronic configuration and,
since the high spin state is the most stable spin case for
[Fe(L1)]2+(d6), a spin crossing from 3IV(3V) to 5IV(5V) arises.
5
6
7
8
9
AUTHOR INFORMATION
Corresponding Authors
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
josepm.luis@udg.edu; miquel.sola@udg.edu; miquel.costas@
udg.edu; perez@dqcm.uhu.es; mmdiaz@dqcm.uhu.es.
5
The final steps from the norcaradiene intermediate, IV, to
5
5
addition products, VI, or from the enol intermediate, III, to
5
Present Addresses
insertion products, V, imply low energy barriers. The ring-
expansion reaction that gives the final cycloheptatriene prod-
uct from norcaradiene intermediate involves a 2.6 kcal/mol
energy barrier, TSIV_VI. On the other hand, the enol-ester
† Institut de Química Computacional i Catàlisi (IQCC) and De-
partament de Química, Universitat de Girona, Campus Montilivi,
17003 Girona (Spain).
5
§ Laboratorio de Catálisis Homogénea, Unidad Asociada al
CSIC, CIQSO-Centro de Investigación en Química Sostenible
and Departamento de Química, Universidad de Huelva, Campus
de El Carmen s/n, 21007 Huelva (Spain).
tautomerism involves a 10.0 kcal/mol energy barrier when it
is assisted by two water molecules, 5TSIII_V. It is interesting to
note that the water is essential to decrease the value of
5TSIII_V, since the equivalent transition state Gibbs energy
barrier without water has a value of ca. 46.6 kcal/mol (see
Table S9 in the SI). Although water was not added in our
experiments, the presence of adventitious water molecules
was experimentally detected by the formation of certain
amount of ethyl glycolate products.12
Author Contributions
‡ These authors equally contributed to this work.
Funding Sources
Support for this work was provided by the MINECO (CTQ2014-
Most importantly, it must be noted that this mechanism ac-
counts for the results on the isotopic labelling experiments; in
first place, the hydrogen atom at the C-H bond that suffers the
attack of the carbene moiety can end up at the benzylic posi-
tion of the product. In addition, the intermediacy of the enol
provides a path for incorporation of hydrogen atoms from
water in the final product.
62234-EXP,
CTQ2014-52769-C3-R-1,
CTQ2014-54306-P,
CTQ2014-52525-P, and grant No. BES-2012-052801 to V.P.),
the Junta de Andalucía (P10-FQM-06292) and the Generalitat de
Catalunya (Project 2014SGR931, Xarxa de Referència en Quími-
ca Teòrica i Computacional, ICREA Academia prizes 2013 to
M.C. and 2014 to M.S.). The EU under the FEDER grant
UNGI10-4E-801 (European Fund for Regional Development) has
also funded this research.
CONCLUSIONS
REFERENCES
The mechanism of the recently discovered iron-catalyzed
selective functionalization of benzene upon carbene insertion
from ethyl diazoacetate has been elucidated. The catalyst
precursor [Fe(OTf)2(L1)] originates an electrophilic iron
carbene intermediate that attacks the arene rendering a
Wheland type intermediate that can engage in cationic rear-
rangements. Experimental data supports an electrophilic aro-
matic substitution behavior. A complementary computational
analysis has demonstrated that the exquisite chemoselectivity
towards the insertion product is rooted in the evolution of this
intermediate via a facile hydrogen atom transfer from the
arene to the carbonyl moiety of the carbene, forming an enol
intermediate. DFT calculations show that the entropy is the
driving force that induce the chemoselectivity of this catalyst
for aryl C-H insertion. Isotopic labelling analysis also provide
credibility to this mechanism. Overall, the current study pro-
vides a mechanistic basis for understanding the unique
chemoselectivity exhibited by this type of iron catalysts in
carbene transfer reactions.
(1) (a) Cheng, Q.-Q.; Doyle, M. P. Adv. Organomet. Chem. 2016,
66, 1-31. (b) Thumar, N. J.; Wei, Q. H.; Hu, W. H. Adv. Organomet.
Chem. 2016, 66, 33-91; (c) Caballero, A.; Díaz-Requejo, M. M.;
Fructos, M. R.,; Olmos, A.; Urbano, J.; Pérez, P. J. Dalton Trans.
2015, 44, 20295-20307; (d) Fructos, M. R.; Díaz-Requejo, M. M.; Pérez, P.
J. Chem Commun 2016, 52, 7326-7335.
(2) Buchner, E.; Curtius, T. Beriche der Dtsch. Chem. Gesellschaft
1885, 18, 2371–2377.
(3) Anciaux, A. J.; Demonceau, A.¸ Noels, A. F.; Hubert, A. J.;
Warin, R.; Teyssie, P. H. J. Org. Chem. 1981, 46, 873-876.
(4) Given that the product obtained corresponds with that in which
the carbene group has been inserted into the C-H bond, we refer to
this as the insertion product. However, the mechanism of this
transformation does not necessarily takes place by an insertion step.
(5) (a) Rosenfeld, M. J.; Shankar, B. K. R.; Shechter, H. J. Org.
Chem. 1988, 53, 2699–2705. (b) Yang, M.; Webb, T. R.; Livant, P. J.
Org. Chem. 2001, 66, 4945–4949.
(6) Fructos, M. R.; Besora, M.; Braga, A. A. C.; Díaz-Requejo, M.
M.; Maseras, F.; Pérez, P. J. Organometallics 2017, 36, 172–179.
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 12
1
2
3
4
(7) (a) Fructos, M. R.; Belderrain, T. R.; de Frémont, P.; Scott, N.
Champouret, Y.; Martín, C.; Álvarez, E.; Etienne, E.; Belderraín, T.
R.; Pérez, P. J. Chem. Eur. J. 2015, 21, 9769-9775.
(19) (a) Khade, R. L.; Zhang, Y. J. Am. Chem. Soc. 2015, 137,
7560–7563. (b) Sharon, D. A.; Mallick, D.; Wang, B.; Shaik, S. J.
Am. Chem. Soc. 2016, 138, 9597–9610
M.; Nolan, S. P.; Díaz-Requejo, M. M.; Pérez, P. J. Angew. Chem.
Int. Ed. 2005, 44, 5284–5288. (b) Rivilla, I.; Gómez-Emeterio, B. P.;
Fructos, M. R.; Díaz-Requejo, M. M.; Pérez, P. J. Organometallics
2011, 30, 2855–2860.
5
6
7
8
9
(8) (a) Liu, Y.; Yu, Z.; Luo, Z.-L.; Zhang, J. Z.; Liu, L.; Xia, F. J.
Phys. Chem. A 2016, 120, 1925−1932. (b) Yu, Z.; Ma, B.; Chen, M.;
Wu, H.-H.; Liu, L.; Zhang, J. J. Am. Chem. Soc. 2014, 136, 6904–
6907. (c)Xi, Y.; Su, Y.; Yu, Z.; Dong, B.; McClain, E. J.; Lan, Y.;
Shi, X. Angew. Chem. Int. Ed. 2014, 53, 9817–9821
(9) (a) Callot, H. J.; Metz, F. Nov. J. Chem. 1985, 9, 167-171. (b)
Davies, H. M. L.; Jin, Q. Tetrahedron Asymmetry 2003, 14, 941-949.
(c) Davies, H. M. L.; Jin, Q.; Ren, P.; Kovalevsky, A. Y. J. Org.
Chem. 2002, 67, 4165-4169. (d) Morilla, M. E.; Díaz-Requejo, M.
M.; Belderrain, T. R.; Nicasio, M. C.; Trofimenko, S.; Pérez, P. J.
Organometallics 2004, 23, 293–295.
(10) (a) Osako, T.; Panichakul, D.; Uozumi, Y. Org. Lett. 2012,
14, 194–197. (b) Chen, C.; Zhu, S.-F.; Liu, B.; Wang, L.-X.; Zhou,
Q.-L. J. Am. Chem. Soc. 2007, 129, 12616–12617. (c) Xie, X.-L.;
Zhu, S.-F.; Guo, J.-X.; Cai, Y.; Zhou, Q.-L. Angew. Chem. Int. Ed.
2014, 53, 2978–2981. (d) Miura, T.; Zhao, Z.; Murakami, M. Angew.
Chem. Int. Ed. 2017, 56, 16645-16649.
(11) Rivilla, I.; Sameera, W. M. C.; Alvarez, E.; Díaz-Requejo, M.
M.; Maseras, F.; Pérez, P. J. Dalton Trans. 2013, 42, 4132–4138.
(12) Conde, A.; Sabenya, G.; Rodríguez, M.; Postils, V.; Luis, J.
M.; Díaz-Requejo, M. M.; Costas, M.; Pérez, P. J. Angew. Chem.
2016, 128, 6640–6644.
(13) Codolà, Z.; Lloret-Fillol, J.; Costas, M. Prog. Inorg. Chem.
2014, 59, 447-532.
(14) (a) For trispyrazolylborate ligands see reference 1c. (b) For
trispyrazolylamines see: Haldón, E.; Álvarez, E.; Nicasio, M. C.;
Pérez, P. J. Chem. Commun., 2014, 50, 8978-8981.
(15) Company, A.; Gómez, L.; Güell, M.; Ribas, X.; Luis, J. M.;
Que, L., Jr; Costas, M. J. Am. Chem. Soc. 2007, 129, 15766-15767.
(16) Kulkarni, N. V.; Dash, C.; Jayaratna, N. B.; Ridlen, S. G.;
Khani, S. K.; Das, A.; Kou, S.; Yousufuddin, M.; Cundari, T. R.;
Dias, H. V. R. Inorg. Chem. 2015, 54, 11043-11045.
(17) (a) Doyle, M. P.; Duffy, R.; Ratnikov, M.; Zhou, L. Chem.
Rev. 2010, 110, 704−724. (b) Doyle, M. P.; McKervey, M. A.; Ye, T.
Modern Catalytic Methods for Organic Synthesis with Diazo
Compounds; Wiley: New York, 1998. (c) Davies, H. M. L.; Dick, A.
R. Top. Curr. Chem. 2009, 292, 303−345. (d) Davies, H. M. L.;
Manning, J. R. Nature 2008, 451, 417−424.
(20) (a) Liu, Y.; Xu, W.; Zhang, J.; Fuller, W.; Schulz, C. E.; Li, J.
J. Am. Chem. Soc. 2017, 139, 5023-5026. (b) Khade, R. L.; Fan, W.;
Ling, Y.; Yang, L.; Oldfield, E.; Zhang, Y. Angew. Chem. Int. Ed.
2014, 53, 7574-7578. (c) Li, Y.; Huang, J.-S.; Zhou, Z.-Y.; Che, C.-
M.; You, X.-Z. J. Am. Chem. Soc. 2002, 124, 13185-13193.
(21) Weinhold, F.; Landis, C. R. Valency and bonding: a natural
bond orbital donor-acceptor perspective; Cambridge University
Press: Cambridge, U.K., 2005.
(22) (a) Ramos-Cordoba, E.; Postils, V.; Salvador, P. J. Chem.
Theory Comput. 2015, 11, 1501–1508. (b) Skara, G.; Gimferrer, M.;
De Proft, F.; Salvador, P.; Pinter, B. Inorg. Chem. 2016, 55, 2185–
2199.
(23) Johnson, E. R.; Keinan, S.; Mori-Sánchez, P.; Contreras-
García, J.; Cohen. A. J.; Yang, W. J. Am. Chem. Soc., 2010, 132,
6498-6506.
(24) Lugan, N.; Fernández, I.; Brousses, R.; Valyaev, D. A.;
Lavigne, G.; Ustynyuk, N. A.; Antipin, M. Y.; Ustynyuk, N. A.
Dalton Trans. 2013, 42, 898–901
(25) (a) Padwa, A.; Austin, D. J.; Price, A. T.; Semones, M. A.;
Doyle, M. P.; Protopopova, M. N.; Winchester, W. R.; Tran, A. J.
Am. Chem. Soc. 1993, 115, 8669–8680. (b) Liu, Y.; Yu, Z.; Zhang, J.
Z.; Liu, L.; Xia, F.; Zhang, J. Chem. Sci. 2016, 7, 1988–1995.
(26) (a) Braga, A. A. C.; Maseras, F.; Urbano, J.; Caballero, A.;
Díaz-Requejo, M. M.; Pérez, P. J. Organometallics 2006, 25, 5292–
5300. (b) Nakamura, E.; Yoshikai, N.; Yamanaka, M. J. Am. Chem.
Soc. 2002, 124, 7181–7192. (b) Boultadakis-Arapinis, M.; Gandon,
V.; Prost, E.; Micouin, L.; Lecourt, T. Adv. Synth. Catal. 2014, 356,
2493-2505.
(27) (a) Yamaguchi, K.; Jensen, F.; Dorigo, A.; Houk, K. N.
Chem. Phys. Lett. 1988, 149, 537–542. (b) Wittbrodt, J. M.; Schlegel,
H. B. J. Chem. Phys. 1996, 105, 6574–6577.
(28) (a) Jensen, K. P.; Ryde, U. J. Phys. Chem. A 2003, 107, 7539–
7545. (b) Kuta, J.; Patchkovskii, S.; Zgierski, M. Z.; Kozlowski, P.
M. J. Comput. Chem. 2006, 27, 1429–1437. (c) Qi, X.-J.; Li, Z.; Fu,
Y.; Guo, Q.-X.; Liu, L. Organometallics 2008, 27, 2688–2698.
(29) (a) Swart, M.; Groenhof, A. R.; Ehlers, A. W.; Lammertsma,
K. J. Phys. Chem. A 2004, 108, 5479–5483. (b) Güell, M.; Solà, M.;
Swart, M. Polyhedron 2010, 29, 84–93.
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
(30) (a) Cohen, A. J.; Mori-Sánchez, P.; Yang, W. Chem. Rev.
2012, 112, 289–320. (b) Mangiatordi, G. F.; Brémond, E.; Adamo, C.
J. Chem. Theory Comput. 2012, 8, 3082–3088.
(18) (a) Dzik, W. I.; Xu, X.; Zhang, X. P.; Reek, J. N. H.; de
Bruin, B. J. Am. Chem. Soc. 2010, 132, 10891–10902. (b) Pereira, A.;
10
ACS Paragon Plus Environment

Page 11 of 12
ACS Catalysis
1
2
3
4
5
H(H)
H
N
C
F
C
H
+
EDA + NaBAr4
X
N
CO2Et
+
Fe
CO2Et
X
overnight, 80∫C
N
< 1%
Addition
> 99%
N
Insertion
6
7
8
9
HHO
H
O
C
L Fe
C
H2O
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
H
OEt
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 12
H(H)
H
CO2Et
H
C
N
F
C
+
EDA + NaBAr4
X
X
N
N
CO2Et
+
Fe
1
overnight, 80˚C
< 1%
> 99%
2
3
N
Addition
Insertion
4
5
6
7
8
9
10
11
HHO
H
O
C
L Fe
C
ACS Paragon Plus Envir2onment
H O
H
OEt