Electrophilic Iron Catalyst Paired with a Lithium Cation Enables Selective Functionalization of Non-Activated Aliphatic C?H Bonds via Metallocarbene Intermediates

Article abstract of DOI:10.1002/anie.201905986

Combining an electrophilic iron complex [Fe(^Fpda)(THF)]₂ (3) [^Fpda=N,N′-bis(pentafluorophenyl)-o-phenylenediamide] with the pre-activation of α-alkyl-substituted α-diazoesters reagents by LiAl(OR^F)₄ [OR^F=(OC(CF₃)₃] provides unprecedented access to selective iron-catalyzed intramolecular functionalization of strong alkyl C(sp³)?H bonds. Reactions occur at 25 °C via α-alkyl-metallocarbene intermediates, and with activity/selectivity levels similar to those of rhodium carboxylate catalysts. Mechanistic investigations reveal a crucial role of the lithium cation in the rate-determining formation of the electrophilic iron-carbene intermediate, which then proceeds by concerted insertion into the C?H bond.

Full text of DOI:10.1002/anie.201905986

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Electrophilic iron catalyst paired with a lithium cation enables
selective functionalization of non-activated aliphatic C-H bonds
via metallocarbene intermediates
Authors: Miquel Costas, Alberto Gómez-Herran, Monica Rodriguez,
and Teodor Parella
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201905986
Angew. Chem. 10.1002/ange.201905986
Link to VoR: http://dx.doi.org/10.1002/anie.201905986
http://dx.doi.org/10.1002/ange.201905986

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
Electrophilic iron catalyst paired with a lithium cation enables
selective functionalization of non-activated aliphatic C-H bonds via
metallocarbene intermediates
Alberto Hernán-Gómez,*[a] Mònica Rodríguez,[a] Teodor Parella,[b] Miquel Costas*[a]
Abstract: Combining an electrophilic iron complex [Fe(Fpda)(THF)]2
a) Previous work
(3)
[Fpda
=
N,N’-bis(pentafluorophenyl)-o-phenylenediamide]
Stoichiometric reactions
a.1
compound with pre-activation of -alkyl substituted -diazoesters
reagents by the Lewis acid LiAl(ORF)4 [ORF= (OC(CF3)3] provides
unprecedented access to selective iron catalyzed intramolecular
functionalization of strong alkyl C(sp3)-H bonds. Reactions occur at
25°C, via -alkyl-metallocarbene intermediates, and with
activity/selectivity levels similar to rhodium carboxylate catalysts.
Mechanistic investigations reveal a crucial role of the lithium cation in
the rate determining formation of the electrophilic iron-carbene
intermediate, which then proceeds via concerted insertion into the
C-H bond.
R1
C6F5
C6F5
N
N
Fe
N
N
C6F5
R-CH(Ph)R1
C6F5
R-H
60ºC-80ºC, 15-36h
R-H = Cyclohexene, Cumene and THF
R1 = Ph, CO2Et
2%
C6H5
C6H5
N
Catalytic regime
Cl
N
Fe
N
O
N
C6H5
a.2
O
O
C6H5
O
R2
N2
R2-H
+
80ºC, 24-48h
R2-H = Cyclohexane, THF
Cl
Introduction
10%
N
N
N
Carbon-carbon C(sp3)-C(sp3) bond formation methodologies are
fundamental in many synthetic transformations. Among the
different existing strategies,[1] the formal metal mediated insertion
of carbene fragments into C(sp3)-H bonds[2] is an appealing
methodology due to its greater atom economy when compared
with the most extended cross-coupling procedures, employing
a.3
N
N
Fe
CO2Et
O
N
N
O
S
O
N
EtO2C
R3
N2
S
O
O
R3
O
45ºC. 2-12h, CH2Cl2, NaBArF
R3 = Vinyl, aryl
BArF = tetrakis[3,5-bis-(trifluoromethyl)phenyl]borate
a.4
CO2Et
N2
CO2Et
R4
organometallic coupling partners (boron, zinc or Grignard
R4
Heme Protein
b]
reagents).[1a,
Despite the robustness of C(sp3)-H bonds,[3]
O
25ºC, CH2Cl2, 17h
O
R4 = OMe, Vinyl
making difficult their chemical manipulation, a variety of transition
metal carbene species (copper, cobalt, silver, palladium, rhodium,
and ruthenium), generated from diazo reagents, are capable of
catalyzing these insertion reactions.[4] In contrast, iron systems,
which are attractive from a sustainability perspective,[5] have
shown modest activity being limited to less demanding processes,
including cyclopropanation reactions,[6] and alkylation of
C(sp2)-H,[7] heteroatom –hydrogen,[8] and relatively weak allylic
and benzylic (BDE < 90 Kcal/mol) C(sp3)-H bonds (Scheme 1).[9]
Pioneer studies by Che[10] proved the feasibility of iron (III)
porphyrins to effect this methodology by stoichiometric
functionalization of cyclohexene, cumene and tetrahydrofuran,
using diphenyldiazomethane as carbene precursor at high
temperatures (60°C - 80°C) (Scheme 1 a.1).
b) This work
2.5% [Fe(Fpda)THF]2
25 % LiAl(ORF)4
O
N2
O
O
R5
25ºC, CH2Cl2, 24h
O
R5
R5 = Aryl, alkyl
Scheme 1. State of art of iron-carbene mediating C(sp3)-H functionalization.
Updating this strategy to a catalytic version, Woo[11] reported
intermolecular insertion of carbene fragments, generated from
methyl phenyldiazoacetate, into the C-H bonds of cyclohexane
and tetrahydrofuran, mediated by an iron (III) porphyrin catalyst at
80°C (Scheme 1 a.2). Although this study opened a new ground
for iron chemistry, it was later reported that these diazoester
compounds undergo formal carbene insertion in the absence of
iron, although in lower yields.[12] More recently, intramolecular
alkylation of weak allylic and benzylic C(sp3)-H bonds has been
reported using a heme-like iron (III) phtalocyanine catalysts, and
intermolecular enantioselective alkylation of activated C(sp3)-H
bonds (benzylic and adjacent to heteroatom) has been described
by laboratory-evolved P-450 enzymes (Scheme 1 a.3 and a.4).[9]
This scenario contrasts to the one operating for the isoelectronic
iron oxos,[13] and iron nitrenes[14] in which upon ligand design,
functionalization of less reactive C(sp3)-H bonds (BDE C-H > 90
Kcal/mol) is well stablished.[3] We hypothesized that replacing the
porphyrin (or related), by ligands of highly electrophilic character
will led to formation of electron-poor iron-carbene species by
[a]
[b]
A. Hernán-Gómez, M. Rodríguez, M. Costas
Departament de Química
Universitat de Girona, Institut de Química Computacional i Catàlisi
(IQCC)
C/ M. Aurèlia Capmany 69, 17003 Girona, Catalonia (Spain)
E-mail: miquel.costas@udg.edu
T. Parella
Servei de Resonància Magnètica Nuclear
Universitat Autònoma de Barcelona,
E-08193 Bellaterra, Barcelona, Spain
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
reaction with diazo compounds, being able to overcome
energetically demanding C-H alkylation processes under mild
conditions.
temperature led to the formation of [Fe(Fpda)(thf)]2 (3), as a pale
orange solid in 82% yield (Scheme 2). Similarly, reaction of ligand
2 and {Fe[N(SiMe3)2]2(thf)} using toluene as a solvent media,
resulted in isolation of a dark purple solid identified as compound
[Fe(Mespda)(toluene)] (4) (75% yield) (Scheme 2). Single crystal
X-ray analysis of compound 3 (Figure 1a, Table 1) reveals a
dimeric structure in the solid state formed by two units of
[Fe(Fpda)(thf)] and with a central {Fe2N2} core, where the iron
centers are connected by two bridging nitrogen atoms of two Fpda
ligands. Fulfilling their coordination sphere the metallic centers
are terminally coordinated by the second nitrogen atom of the
N,N’-bidentate ligands, and by a molecule of THF. Inspection of
the monomeric asymmetric unit [Fe(Fpda)(thf)] reveals two
significantly different Fe-N bond distances, Fe1-N1 2.112 (1) and
Fe1-N2 1.969 (1), attributed to the bridging character of N1
between Fe1 and Fe1’, while N2 coordinates terminally to Fe1.
Similarly, within the {Fe2N2} core two different Fe-N bond
distances are observed, being the previous Fe1-N1 bond distance
longer than the one formed between the same N1 and the
neighboring iron atom Fe1’-N1 2.069(1). These Fe-N bond
distances are within the range of high spin Fe(II) compounds
supported by bidentate ligands (Fe-N> 2.0 Å).[17] In contrast,
compound 4 (Figure 1b, Table 1) exhibits a monomeric 6-toluene
adduct with Mespda acting as a terminal N,N’-bidentate ligand and
with shorter Fe-N bonds (average 1.87(1) Å).
In our aim to develop iron-carbene species with the ability to
functionalize strong C(sp3)-H bonds, herein we report an
electrophilic iron complex [Fe(Fpda)(THF)]2 (3) [Fpda = N,N’-
bis(pentafluorophenyl)-o-phenylenediamide], which is arranged
in an unique dimeric fashion within ortho-phenylenediamide iron
chemistry, which traditionally has been dominated by formation of
the more stable [FeL2] fragment, as described by Wieghardt.[15]
For comparison reasons, the related complex [Fe(Mespda)(Tol)]
(4)
[
Mespda
=
N,N’-bis(2,4,6-trimethylphenyl)-o-
phenylenediamine], displaying poorer electrophilic nature, has
been also investigated. Effectively, we disclose unprecedented
iron catalyzed intramolecular carbene insertion into a variety of
non-activated aliphatic C-H bonds, via iron-carbene intermediates
(Scheme 1b), under mild conditions (25ºC). The reaction is
critically based on two novel aspects; a) pre-activation of the
diazoester reagent by the Lewis acid LiAl(ORF)4 [ORF=
(OC(CF3)3], which enables formation of the iron-carbene species
under mild conditions, and b) the electrophilic character of
complex 3, which permits functionalization of non-activated
aliphatic C-H bonds. Remarkably 3 exhibits activity and selectivity
levels comparable to rhodium carboxylate catalysts.
Results and Discussion
Catalysts
phenylenediamine ligands
nucleophilic aromatic
synthesis
and
1
characterization.
were synthesized by
between lithium
Ortho-
and
2
substitution
1,2-phenylendiamide and hexafluorobenzene for compound 1,
and Buchwald-Hartwig amination of 1,2-dibromobenzene with
2,4,6-trimethylaniline in the case of 2, according to previous
reports in the literature.[16]
Figure 1. Molecular structures of a) compound 3 and b) compound 4 with
thermal ellipsoids at 50% probability. Hydrogen atoms are omitted for clarity.
Table1. Selected bond distances of compounds 3 and 4.
Bond distances / Angles
Fe1-O1
3
4
-
2.017(1)
2.112(1)
2.069(1)
1.969(1)
-
1.446(2)
1.399(2)
1.411(2)
1.390(3),
1.399(3),
1.388(3)
1.380(3),
1.392(3)
Fe1-N1
1.881(2)
-
Fe1-N’1
Fe1-N2
Fe-CTol average
N1-C1
N2-C2
C1-C2, ,
C1-C6
C2-C3
C3-C4
C4-C5
1.867 (2)
2.09(1)
1.359(4),
1.359(4),
1.426(4),
1.410(4),
1.414(4),
1.375(4),
1.407(4),
1.382(4);
C5-C6
Scheme 2. Synthesis of compounds 3 and 4.
The different nature of the formed complexes with the
non-nitrogen donor ligands, THF -adduct in 3 and toluene
-adduct in 4, points to different electronic configuration of each
iron complex due to the “redox-active” nature of the N,N’-
Metallation of
1
with the bisamide iron compound
{Fe[N(SiMe3)2]2(thf)} in a mixture THF/hexane (1:9) at room
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
bidentate employed ligands. The latter is confirmed by
comparison of the metrical data of the ligand backbones of 3 and
4, following Wieghardt´s structural approach to determine the
redox form of the ligand and metal in complexes supported by
ortho-phenylene diamine ligands.[15] Thus compound 3 displays
an aromatic phenylene ring (C-Caverage 1.39(1) Å) and single C-N
bonds (C-Naverage 1.42(2) Å), whereas species 4 exhibits subtle
dearomatization on the phenylene ring with two sets of C-C bonds
(average C-C and C-C1.417(8) Å; C-C average 1.378(5) Å), as
well as shorter C-N bond distances (C-Naverage 1.359(4) Å). These
data are consistent with a dianionic ortho-phenylenediamide
bidentate ligand binding an Fe2+ metallic center in complex 3,
similarly to the homoleptic [Fe(Fpda)2][AsPh4] previously reported
by Wieghardt, although in this case two Fpda2- coordinate the
same metallic center.[15b] Key for the isolation of compound 3
relies on the use of the iron base {Fe[N(SiMe3)2]2(thf)}, which
precludes the presence of an excess of dianionic diamide ligand,
as when the lithiated diamide is used as transmetallating reagent,
and ultimately avoids formation of the [FeL2] fragment.
unsaturated ester side product 5b, generated in a -hydride
migration process, or alternatively by two consecutive hydrogen
atom abstraction (HAT) steps, as it has been recently reported by
De Bruin for intramolecular C-H alkylation catalysis via radical
cobalt-carbene intermediates. [19]
Table 2. Optimization for catalytic functionalization of diazo reagent 5.a
In contrast, compound 4 is better described as an iron (I) ortho-
diiminosemiquinonate with metrical data similar to the
[Fe(Iprpda)(Tol)] compound reported by Song.[18] In agreement
with the electronic structures suggested by X-ray crystallography,
1H NMR characterization in C6D6 solution displayed different
magnetic behavior for the two complexes. Compound 4 is
diamagnetic and its spectrum displays six signals in the range 7
– 2 ppm, assigned to the N,N’-chelating ligand, and four high
shielded resonances (5.14, 4.89 and 4.82 and 1.78 ppm)
consistent with a 6-toluene adduct. The observed diamagnetic
behavior of
4
can be rationalized in terms of an
antiferromagnetically coupling between the structurally deduced
FeI (d7) metallic center with the unpaired electron on the
ortho-diiminosemiquinonate ligand, as it was reported for the
isostructural [Fe(Iprpda)(Tol)].[18] In contrast, in C6D6 species 3
displays a paramagnetic behavior showing two resonances at 25
and 15 ppm, accounting for the four aromatic protons at the
backbone of the bisamide ligand, and a solution magnetic
moment eff = 4.71 (3) B per dimer at room temperature.
Confirming retention of the solid state structure of 3 in C6D6
solution, when its 1H NMR spectrum was registered in [D8]-THF,
a shift of the latter signals, assigned to the dimer (25 and 15 ppm),
was observed towards higher field (-18 and -22 ppm), in
agreement with THF coordination to the iron centers and breaking
up of the dimer to generate a monomer species. Thus, the
deduced spin-only value for compound 3 in C6D6, which is lower
than the expected for two uncoupled high-spin iron (II) (6.9 B),
can be attributed to antiferromagnetic coupling between the two
vicinal iron atoms (2.615(1) Å) of the dimer.
a Reaction conditions: [5] = 0.08 mM. Yields of 5a, 5b, and diastereomeric ratios
were determined by 1H NMR spectroscopy using dibromomethane as internal
standard.
Despite this lack of reactivity, during the addition of diazo
compound 5 over the CH2Cl2 solution of 3 we observed an
immediately color change from orange to dark brown. Analysis of
the mixture resulting from the reaction of stoichiometric amounts
of diazoester 5 and iron species 3 by infrared spectroscopy shows
a band shift to a lower value ( = 13 cm-1) of the C=O functional
group (Figures 2a-2c), while the band assigned to the N=N
functional group is modified in a more modest extend ( = -8
cm-1). These observations suggest that the iron compound is
unable to effect formation of the metallocarbene intermediate
upon nitrogen release of the diazo compound, and instead a novel
adduct species through the binding of the O donor atom of the
carbonyl group [Fe-O(C)] present in the diazo reagent to the
Lewis acid FeII atom of 3 is formed. This scenario is reminiscent
to the one recently reported for the C(sp2)-H bond alkylation
reactions via non heme iron-carbene species. In this case,
computational studies rationalize the high reaction temperatures
(80°C) required by dissociation of an unproductive Fe-O(C)
adduct formed between the iron catalyst and the oxygen atom of
the carbonyl group of ethyl diazoacetate.[7a]
Catalytic carbene transfer reactions. To begin our
investigations on C-H alkylation, we tested the catalytic activity of
compounds 3 (2.5 mol%) and 4 (5 mol%) in the intramolecular
C-H alkylation of diazoester 5 in CH2Cl2 at 25°C for 24 hours
(Table 1, entries 1-2). The reactions provided poor conversions
(7-8%) of 5 into the desired 5a product, where intramolecular
alkylation has taken place at the benzylic position, and the -
Lithium activation of the azo reagent. We hypothesized that the
undesired Fe-O(C) interaction can be disfavored by addition of a
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
stronger Lewis acid than the iron centers present in compounds 3
and 4.
resulted in enhanced conversions of 5 at 25°C (Table 2, entries
3-4), exhibiting 3 the highest reactivity and selectivity with a 48%
conversion and a 72:28 ratio (5a:5b), while 4 only rendered a 18%
conversion with a 47:53 ratio (5a:5b). Seeking to increase the
yield of product 5a we optimized the loading of lithium salt
LiAl(ORF)4 (Table 2, entries 5-9), resulting in nearly quantitative
conversion of 5, the best 5a/5b ratio, and a high 5a diastereomeric
ratio (76/24, syn:anti) at 25 mol% concentration (Table 2, entry 6).
Of notice this lithium concentration actually matches the absence
of free diazo compound observed by infrared studies. Proving the
synergistic partnership between Fe and Li, when the lithium
aluminate LiAl(ORF)4 is reacted with diazo compound 5 at 25 °C
in the absence of catalyst 3, no 5a or 5b product was observed
(Table 2, entry 10). Under the optimized conditions we
investigated the influence of each component in the catalytic
reaction. Other alkali metal salts, such as NaBArF, have been
used in more facile C(sp3)-H insertion reactions via iron-carbene
species, playing a mere role as trapping reagents of the halogen
Figure 2. Expanded IR spectra displaying C=O and N2 bands of a) diazoester
5, b) reaction mixture of 5 and 3 (1:0.5), c) reaction mixture of 5 and 3 (1:1), d)
reaction mixture of 5 and LiAl(ORF)4 (1:0.05), e) reaction mixture of 5 and
LiAl(ORF)4 (1:0.15) and f) reaction mixture of 5 and LiAl(ORF)4 (1:0.25). * band
belonging to compound 3. For full spectra see supporting information
atoms connected to the iron center, and generating a cationic and
9b]
hence more reactive iron species.[7,
In our case replacing
Indeed, repeating the infrared spectroscopy monitoring of the
mixture formed by compound 5 and variable amounts of the
lithium salt LiAl(ORF)4 [ORF= (OC(CF3)3] (5-25 mol%), we
observed a split in two of the investigated C=O bands, assigned
to free 5 (1690 cm-1) and lithium bound 5 (1654 cm-1) (Figures 2d-
2f). According to this assignation, the intensity of the free diazo
band systematically decreased to depletion in favor of the adduct
band as the concentration range of LiAl(ORF)4 increased from 5
to 25 mol%. Beyond 25 mol% concentration of lithium aluminate
no changes were observed in the IR spectra. Of relevance to
these observations, this coordination mode reproduces the one
observed for the interaction of lactones with lithium inorganic salts,
which occurs through the carbonyl group.[20] In parallel, within the
range of LiAl(ORF)4 concentrations studied, the N=N stretching
frequency experiences a 15 cm-1 shift to higher energies, which
can be rationalized by an additional, but weaker, interaction
between a lithium cation and 5 through the negatively polarized
carbon bearing the diazo fragment Li-C(N2). In solution, a
dynamic binding of lithium to the diazoester 5 can be deduced by
1H NMR study of the previously described reaction mixtures within
the concentration range 5-25 mol% of LiAl(ORF)4, which display
only one set of broad signals of the diazoester reagent (Figures
S127-S128, supporting information). Further evidence for the
strong binding of the diazo compound 5 to lithium is provided by
mass spectroscopy analysis of a reaction mixture of 5 and
LiAl(ORF)4 (1:0.25), which exhibits peaks at m/z = 471.2585 and
239.1384, corresponding to a mono- and bis-adduct between 5
and the lithium cation (Figures S129-S131, supporting
information). Collectively, the data suggests the formation of a
strong adduct between 5 and lithium, through the carbonyl group,
which may be used to liberate the iron center, setting it ready for
carbene formation.
LiAl(ORF)4 by NaBArF (Table 2, entry 11) turned into lower
reactivity observing only 7% yield of 5a, proving sodium to be a
less affective co-catalyst in agreement with the softer nature as a
Lewis acid compared to lithium cation.
Highlighting the importance of low coordinate iron centers to
develop these C-H insertion reactions, when catalyst 3 was
substituted by an iron complex supported by a polydentate amino-
based ligand [Fe(OTf)2(PyTACN)] (PyTACN
=
1-(2-
pyridylmethyl)-4,7-dimethyl-1,4,7-triazacyclononane, Table 2,
entry 12),[7] no product 5a was formed recovering the starting
material 5. This catalyst has been recently shown to activate ethyl
diazoacetate to effect chemoselective arene alkylation of alkyl
arenes via a formal carbene insertion reaction: Reactions require
80°C to proceed and did not display sp3 alkylation, even in the
presence of weak tertiary C-H bonds. In addition, and for
comparison reasons the studied cyclization was also subjected to
rhodium catalysis,[19] which in our hands using 2.5 mol% of
[Rh(OAc)2]2 provided similar results to our iron-lithium catalytic
system (Table 2, entry 13). Drawing a comparison to iron catalysis
in terms of sustainability, the earth-abundant copper metal
supported by trispyrazolylborate compounds have emerged as
powerful
hydrocarbons,[21] however alkylation of 5 using the electron-
deficient (hydrotris(3,4,5-
TpBr3Cu(NCMe) (TpBr3
catalysts
able
to
functionalize
unreactive
=
tribromo)pyrazolylborate)] species rendered almost exclusively
the undesired 5b product (Table 2, entry 14).
Substrate scope. With the optimized conditions in hand, we next
investigated the viability of this catalytic Fe/Li system in
intramolecular C-H alkylation of diazo substrates with varying
degrees of electrophilicity (6-8, Table 3 ) and containing bonds of
different strength (9-19, Table 3).
Li-promoted carbene transfer reactivity. In agreement with the
above described analysis, repeating the intramolecular C-H
alkylation reaction of 5 catalysed by 2.5 mol% of 3 and 5 mol% of
4, now in presence of 5 mol% of LiAl(ORF)4 [ORF= (OC(CF3)3]
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
Table 3. Substrate scope a
in yields are in agreement with involvement of an electrophilic
iron-carbene as the active species, with enhanced reactivity as
the nucleophilicity of the reactive C-H bond increases. Notably,
directed by the formation of a favorable five-membered ring a
strong alkyl C-H bond is functionalized over the weaker benzylic
C-H bond, forming the five-membered ring 11a, albeit in a modest
yield 30%. These catalytic C-H alkylation reactions compared well
with the only previous report in the literature for the catalytic
intramolecular functionalization of benzylic C-H bonds via iron
carbene intermediates, as the latter required heating (45°C) and
higher catalyst loading (10 mol%) to carry out these reactions
under similar reaction times and yields.[9b] In addition, these
results (Table 3, 5a, 9a-11a) suggest that formation of olefin side
products is substrate-dependent, becoming this reaction more
competent as the functionalization of the desired C-H bond is less
favorable, as observed for substrates 10 and 11 in which the
-unsaturated ester product is generated in higher yields. Of
relevance, reactions exhibit an excellent mass balance; for each
substrate, the sum of yields of alkylated product and the
-unsaturated ester account for the total amount of initial
substrate. Encouraged by the formation of 11a under mild
conditions, we next explored functionalization of strong C-H
bonds found in alkyl derivatives. Starting with a less demanding
one, catalyst 3 in combination with LiAl(ORF)4 was able to produce
cyclization upon functionalization of a tertiary C-H bond present
in diazoester reagent 12, generating 12a in 56% isolated yield. A
similar yield (64%) was obtained for cyclopentane derivative 13a,
when substrate 13 containing a more challenging secondary C-H
bond was subjected to the catalytic Fe/Li system. As was
observed for benzylic C-H bonds, formation of a five-membered
ring primes over the possibility of forming larger size rings as in
the case of diazoester 14, which renders 14a in 60% yield. More
importantly, formation of cyclopentane derivatives 11a-14a
occurs in an excellent diastereomeric ratio (anti:syn ≥ 95:5).
Beyond monocyclic products, the developed methodology gives
access to bicyclic spirocompounds (15a, 16a) and fused ring
products (17a, 18a) in moderate to good yields (47-70%).
Contrastingly, catalyst 3 in combination with the lithium salt
activates the most challenging primary C-H bonds present in
substrate 19 in nearly stoichiometric yield (6%).
Aceptor / Acceptor Donor / Acceptor
N2
Donor / Donor
TsHN
N
N2
3 (2.5 mol%), Li(AlORF)4
25ºC, 24h, CH2Cl2
N2
O
O
O
O
O
6
O
Y
X
7
8
98% rsm
83% rsmb
70% rsmb
O
N2
3 (2.5%), LiAl(ORF)4 (25 mol%)
O
O
R
+
O
R
25ºC, 24h, CH2Cl2
O
6-16
5a, 9a-14a (anti:syn)
5b-19b (E:Z)
O
R
O
O
N2
N2
O
O
O
O
O
O
O
5
5a 60% (76:24) [57% (89:11)]
14
14a 68% (91:9) [60% (96:4)]
14b 30% (23:77)
5b 35% (19:81)
O
N2
N2
O
O
O
O
O
O
MeO
MeO
9
9a 73% (77:23) [70% (79:21)]
15
15a 68% [70%]
15b 31% (26:84)
9b 24% (29:71)
O
N2
N2
O
O
O
O
O
O
O
Cl
Cl
10
10a 53% (77:23) [39% (86:14)]
10b 43% (23:77)
16
16a 71% [63%]
16b 20% (52:48)
O
O
N2
N2
O
O
O
O
O
O
c
11
11a 26% (>99:1) [30% (>99:1)]
11b 71% (23:77)
17
18
17a 56% [47% ]
17b 31% (25:75)
O
O
N2
N2
O
O
O
O
O
O
O
d
12
12a 70% [56%]
18a 76% [67%]
12b 26% (14:86)
18b 10% (29:71)
O
N2
N2
O
O
O
O
O
O
13
13a 59% (95:5) [64% (95:5)]
13b 39% (24:76)
19
19a 6% (>99:1)
19b 93% (25:75)
a Reaction conditions: [5-19] = 0.08 mM. Conversions of 6-8 were determined
by 1H NMR spectroscopy using dibromomethane as internal standard. Yields of
5a, 9a-19a and 5b,9b-19b, were determined by 1H NMR spectroscopy using
dibromomethane as internal standard. Diastereomeric ratios were determined
by GC and 1H NMR spectroscopy. Anti diastereomer for products 5a, 9a-14a
was determined to be the major isomer by NMR characterization. Isolated yields
for 1 mmol scale reactions are reported in brackets and their corresponding
diasteromeric ratios determined by GC and 1H NMR spectroscopy. rsm =
b
c
recovered starting material; Non-identified products; d.r. (63:25:12), d.r. for
isolated yield (68:29:3) determined by GC and 1HNMR spectroscopy, relative
stereochemistry was not determined; d.r (94:6) determined by GC, relative
d
stereochemistry was not determined.
Mechanistic analyses. We envisioned that the reaction will
proceed via a concerted alkylation process (Figure 3a), in which
the iron-carbene is formed by nitrogen extrusion from the
corresponding lithium-diazoester adduct, and then effects
insertion into the C-H bond forming a new C-C and C-H bond
simultaneously, releasing the organic product and re-generating
the active catalyst. Nevertheless, two alternative scenarios can be
proposed, such as formation of free carbene species able to effect
the observed C-H insertion reactions or a hydrogen atom transfer
initiated radical stepwise process, as in the case of iron mediated
C-H oxidation,[13] amination,[14] and a recently described iron-
phthalocyanine catalyzed alkylation.[9b] To distinguish between
these three possibilities, we subjected our system to a variety of
mechanistic analyses. First, generation of a free carbene was
thermally (80°C) achieved, and resulted in exclusive formation of
the undesired product 5b in a 48% yield (Figure 3b). Under the
same conditions (80°C and absence of iron catalyst), addition of
the Lewis acid LiAl(ORF)4, afforded exclusively 5b in quantitative
yields (Figure 3b). This acceleration is in agreement with the
Modification of the electronic properties of the diazo reagent
resulted in no reaction in the case of acceptor/acceptor substrate
6, and in low conversion of the donor/acceptor compound 7
towards a mixture of unidentified compounds. In addition, we also
explored other carbene sources such as N-tosylhydrazone 8
which gives rise to the corresponding diazo reagent in situ upon
deprotonation using a base (LiOtBu), as it has successfully used
in cobalt catalysis.[4d-g] However this methodology proved to be
incompatible with the developed Fe/Li system.
Extending the study to donor/acceptor diazo reagents structurally
similar to 5 and containing benzylic C-H bonds, using the electron-
rich para-methoxy aryl diazoester reagent
9 led to the
corresponding cyclopentane product 9a in a high isolated yield
70%, while the electron-poor para-chloro substituted compound
10 was transformed to 10a in a diminished 39% isolated yield.
Both compounds 9a and 10a are formed in similar and modest
diastereomeric ratio (d.r.) 77:23. Again, the differences observed
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
observed interaction Li-C(N2) by IR spectroscopy, which at high
temperatures favors formation of a free carbene. The lack of
formation of 5a in absence of iron catalyst 3 provides strong
support for the necessary formation of an iron-carbene
intermediate as the active species responsible for the C-H
alkylation reactions.
This experiment revealed two different values for each
diastereomer, being the isotopic effect lower in the case of the
major anti isomer [KIE = 2.3 (1)] than the one observed for the syn
isomer [KIE = 4.3 (1)] (Figure 3c). For comparison reasons
diazoester [D1]-5 was also subjected to catalytic alkylation using
the dimeric [Rh(OAc)2]2 catalysts, well-known to operate through
C-H insertion,[22] which afforded again two KIE values of 2.3 (1)
for the anti isomer, and 3.7 (1) for the syn isomer (Figure 3d). The
similar KIE values for both catalysts support the proposed
concerted mechanism. In addition, the different results for each
diastereomer in both metal-mediated reactions reflect two
different transition states for the concerted insertion, as it has
been also observed in previous rhodium mediated carbene
insertion reactions.[23] To confirm the absence of radical
intermediates, the radical trap TEMPO was added to the
optimized alkylation reaction of diazo reagent 5, however this
experiment led to fully recovery of the starting material (Figure 3d).
This can be reasoned by preliminary reaction between the
catalyst 3 and TEMPO, rendering a species unable to form an
iron-carbene active species. We next explored cyclization of the
radical clock substrate 19, which leads to the cyclopentane 19a in
72% yield, conserving intact the cyclopropyl fragment (Figure 3e).
Hydride shift product 19b, accounts for the rest of the starting
material. The lack of rearranged products is consistent with an
insertion reaction, devoid of long lived radical or carbocationic
intermediates.
a)
C-H insertion
Iron-carbene formation
N2
O
O
H
[Fe(Fpda)(THF)]2
H
H
R
O
O
R
H
Al(ORF)4
Li
or
O
R
Li[Al(ORF)4]
-N2
O
b-hydride migration
[Fe]
H
H
R
O
O
Once the concerted mechanism was established, we explored the
nature of the rate determining step by kinetic studies. H-NMR
O
H/D
O
c)
1
N2
H
D
Cat, 25ºC
O
D/H
spectroscopy monitoring of the C-H alkylation of substrate 5 under
the optimized conditions (2.5 mol% of 3 and 25 mol% of
LiAl(ORF)4) reveals a first order dependence on the diazoester
reagent (Figure S135). In addition, a KIE value of 1 was
determined by product analysis of the competitive reaction
between equimolar amounts of 5 and [D1]-5 at low substrate
conversion (15%) (Figure 3c). Finally, variable time normalization
analysis[24] for the intramolecular insertion reaction at 5, at
different concentrations of LiAl(ORF)4) (25-40mol%) revealed a
fractional 1.5 order (Figure S136), suggesting participation of
more than one lithium cation through a coordination-dissociation
equilibrium process. This is in agreement with the IR studies,
which show that 5 interacts with lithium through the carbonyl
group Li-O(C), and the carbon negatively polarized Li-C(N2),
requiring the latter to be replaced by iron for the formation of the
iron-carbene species. Overall, these data are consistent with a
rate-limiting iron-carbene formation, preceded by competition
between a second lithium cation and iron for the C(N2) center in
an equilibrium process, shifted towards the interaction with iron
due to the irreversible iron-carbene formation.
O
CH2Cl2, 24h
[D1]-5
[D1]-5a
Intramolecular Competition
Parallel rates
KIE
Catalyst
syn
anti
3 (2.5 mol%)/LiAl(ORF)4 (25 mol%)
4.3 (1)
2.3 (1)
kH/kD = 1
[Rh(OAc)2]2 (2.5 mol%)
3.7 (2)
2.3 (1)
d)
N2
3 ( 2.5 mol%) LiAl(ORF)4 (25 mol%)
TEMPO (1 equiv.) 25ºC, CH2Cl2, 24h
O
96% rsm
O
5
e)
O
N2
24h, 25ºC, CH2Cl2
O
O
O
+
3 (2.5 mol%)
LiAl(ORF)4 (25 mol%)
O
O
20
20a
70% (85:15)
[72% (85:15)]iii
20b
16% (36:64)iv
iii Reaction conditions: [19] = 0.08 mM. Yields determined by 1H NMR spectroscopy using
dibromomethane as internal standard. Diastereomeric ratios (anti:syn) determined by GC;
iv Diastereomeric ratios (E:Z) determined by 1H NMR.
Figure 3. a) Proposed mechanism of reaction; b) Generation of free carbene
via thermal decomposition, c) Plot of Ln[5] vs time (h) for C-H alkylation reaction
of 5 catalyzed by 2.5 mol% of 3 and 25 mol% of LiAl(ORF)4; d) Intramolecular
KIE, e) C-H alkylation using radical probe 17.
Conclusion
In conclusion, we describe an electrophilic iron complex 3 which
in combination with LiAl(ORF)4 as co-catalyst activates
diazoesters under mild reaction conditions, resulting in
intramolecular C(sp3)-H alkylation of a variety of benzyl and alkyl
C-H bonds, leading to a variety of carbocyclic cyclopentanes as
well as bicyclic spiro and fused ring compounds. High
Next, to shed light on the nature of the C-H activation step,
intramolecular kinetic isotope effect (KIE) was explored via
alkylation of the partially deuterated diazoester reagent [D1]-5.
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
Zhang, J. Am. Chem. Soc. 2018, 140, 4792–4796; c) X. Cui, X. Xu, L.-
M. Jin, L. Wojtas, X. P. Zhang, Chem. Sci., 2015, 6, 1219–1224; d) C. te
Grotenhuis,B. G. Das, P. F. Kuijpers, W. Hageman, M. Trouwborst, B.
de Bruin, Chem. Sci., 2017, 8, 8221–8230; e) C.te Grotenhuis, N. van
den Heuvel, J. I. van der Vlugt, B. De Bruin, Angew. Chem. Int. Ed. 2018,
57, 140–151; Angew. Chem. 2018, 130, 146–151; f) A. S. Karns, M.
Goswami, B. de Bruin, Chem. Eur. J. 2018, 24, 5253–5258; g) X. Wen,
Y. Wang, X. P. Zhang, Chem. Sci., 2018, 9, 5082–5086; h) A. Gutierrez
Bonet, F. Julia-Hernandez, B. de Luis, R. Martin, J. Am. Chem. Soc.
2016, 138, 6384−6387; i) A. R. Reddy, C. Y. Zhou, Z. Guo, J. H. Wei, C.
M. Che, Angew. Chem. Int. Ed. 2014, 53, 14175−14180; Angew. Chem.
2014,126, 14399 –14404.
stereoselectivity is observed in the alkylation of the strong
aliphatic sites. Mechanistic studies strongly suggest that the
reactions proceed via highly reactive iron-carbene active species.
Key to its formation is the binding of the carbonyl group of the
diazoester reagent to lithium Li-O(C). Presumably, this reaction
blocks formation of unproductive Fe-O(C) species, favoring
interaction of the diazo moiety with the iron center, previous
replacement of a second lithium cation interacting with the diazo
reagent via Li-C(N), which then triggers release of nitrogen and
formation of the reactive iron-carbene in the rate determining step.
In addition these studies show a close similarities to rhodium
based carbene transfer reactions, which entail a concerted
metallocarbene insertion into the C-H bond, forming the new C-C
and C-H bonds, suggesting that the current iron based system
also operates via a C-H insertion process. Most relevant, the
current systems extend the substrate scope of iron carbene
species towards site selective alkylation of strong aliphatic C-H
bonds, opening transformations so far only accessible with
precious metals. Key to this finding is the tuning of the ligand
architecture to endorse the iron center with a highly electrophilic
character. Considering the structural versatility of this class of
ligands, and the high level of diastereoselectvity observed in the
functionalization of aliphatic sites, the development of catalyst
dependent site and enantioselective transformations, akin those
stablished for rhodium carbenes is envisioned.
[5]
[6]
a) A. Furstner, Acs Central Science 2016, 2, 778-789; b) S. Enthaler, K.
Junge, M. Beller, Angew. Chem. Int. Ed. 2008, 47, 3317-3321; c) I. Bauer,
H. J. Knolker, Chemical Reviews 2015, 115, 3170-3387.
a) P. S. Coelho,E. M. Brustad, A. Kannan, F. H. Arnold, Science 2013,
339, 307−310; b) J.-J. Shen, S.-F. Zhu, Y. Cai, H. Xu, X.-L. Xie, Q.-L.
Zhou, Angew. Chem. Int. Ed. 2014, 53, 13188-13191; Angew. Chem.
2014,126, 13404 –13407
[7]
a) V. Postils, M. Rodríguez, G. Sabenya, A. Conde, M. M. Díaz-Requejo,
P. J. Perez, M. Costas, M. Sola, J. M. Luis, ACS Catal. 2018, 8,
4313−4322; b) A. Conde, G. Sabenya, M. Rodríguez, V. Postils, J. M.
Luis, M. M. Díaz-Requejo, M. Costas, J. P. Pérez, Angew. Chem. Int. Ed.
2016, 55, 6530-6534; Angew.Chem. 2016,128,6640–6644
[8]
[9]
a) G. Sreenilayama, R. Fasan, Chem. Commun. 2015, 51, 1532-1534;
b) L. K. Baumann, H. M. Mbuvi, G. Du, L. K. Woo Organometallics, 2007,
26, 3995–4002; c) E. H. Wang, Y.-J. Ping, Z.-R. Li, H. Qin, Z.-J. Xu, C.-
M. Che, Org. Lett., 2018, 20, 4641–4644.
a) R. K. Zhang, K. Chen, X. Huang, L. Wohlschlager, H. Renata, F. H.
Arnold, Nature, 2019, 565, 67-72; b) J. R. Griffin, C.I. Wendell, J. A.
Garwin, C. M. White, J. Am. Chem. Soc. 2017, 139, 13624−13627.
Acknowledgements
[10] Y. Li, J.-S. Huang, Z.-Y. Zhou, C.-M. Che, X.-Z. You, J. Am. Chem. Soc.
2002, 124, 13185-13193.
[11] H. M. Mbuvi, L. K. Woo, Organometallics 2008, 27, 637–645.
[12] C. Tortoreto, D. Rackl, H. M. Davies, Org. Lett. 2017, 19, 770−773.
[13] a) P. R. Ortiz de Montellano, Chem. Rev. 2010, 110, 932-948; b) S.
Chakrabarty, R. N. Austin, D. Deng, J. T. Groves, J. D. Lipscomb, J. Am.
Chem. Soc. 2007, 129, 3514-3515; c) J. C. Price, E. W. Barr, T. E. Glass,
C. J. Krebs, J. Martin Bollinger, J. Am. Chem . Soc. 2003, 125, 13008 –
13009; d) K. Chen, L. Que, J. Am. Chem. Soc. 2001, 123, 6327-6337; e)
A. Company, L. Gómez, M. Güell, X. Ribas, J. M. Luis, L. Jr. Que, M.
Costas, J. Am. Chem. Soc. 2007, 129, 15766-15767.
We acknowledge the Spanish Ministry of Science CTQ2015-
70795-P (M.C.) and PGC2018-095808-B-I00 (T.P.). AHG
acknowledges the support of the Beatriu de Pinós postdoctoral
programme of the Government of Catalonia's Secretariat for
Universities and Research of the Ministry of Economy and
Knowledge and to the financial support from Marie Curie
Individual Fellowship (H2020-MSCA-IF-2017). We thank Anna
Company, for helpful discussions, and S.T.R. of U.d.G. are
acknowledged for experimental support.
[14] a) E. T. Hennessy, T. A. Betley, Science 2013, 340, 591−595; b) B. Bagh,
D. L. J. Broere, V. Sinha, P. F. Kuijpers, N. P. van Leest, B. de Bruin, S.
Demeshko, M. A. Siegler, J. I. van der Vlugt, J. Am. Chem. Soc. 2017,
139, 5117−5124.
Keywords: Iron • carbene • C-H activation
[15] a) M. M. Khusniyarov, T. Weyhermüller, E. Bill, K. Wieghardt, J. Am.
Chem. Soc. 2009, 131, 1208–1221; b) M. M. Khusniyarov, E. Bill, T.
Weyhermüller, E. Bothe, K. Harms, J. Sundermeyer, K. Wieghardt,
Chem. Eur. J. 2008, 14, 7608 – 7622.
[1]
[2]
For comprehensive reviews in C(sp3)-C(sp3) methodologies see: a) N.
Kambe, T. Iwasakia, J. Terao, Chem. Soc. Rev., 2011, 40, 4937–4947;
b) R. Jana, T. P. Pathak, M. S. Sigman, Chem. Rev. 2011, 111, 1417–
1492; c) Z. Dong, Z. Ren, S. J. Thompson, Y. Xu, G. Dong, Chem. Rev.
2017, 117, 9333−9403.
[16] a) V. M. Jiménez-Pérez, M. Ibarra-Rodríguez, B. M. Muñoz-Flores, A.
Gómez, R. Santillan, E. Hernández Fernadez, S. Bernès, N. Waksman,
R. Ramírez Duron, J. Mol. Struct. 2013, 1031, 168–174; b) M. M.
Khusniyarov, K. Harms, O. Burghaus, J. Sundermeyer, B. Sarkar, W.
Kaim, J. van Slageren, C. Duboc, J. Fiedler, Dalton Trans. 2008, 1355-
1365.
a) M. M. Díaz-Requejo, A. Caballero, M. R. Fructos, P. J. Pérez, Alkane
C-H Activation by Single-Site Metal Catalysis, Springer, Amsterdam,
2012, Chap.6; b) A. Caballero, M. M. Diaz-Requejo, M. R. Fructos, A.
Olmos, J. Urbano, P. J. Perez,. Dalton Trans. 2015, 44, 20295−20307;
c) A. Caballero, P. J. Perez, Chem. Eur. J., 2017, 23, 14389-14393; d)
A. Ford, H. Miel, A. Ring, C. N. Slaterry, A. R. Maguire, M. A. McKervey,
Chem. Rev. 2015, 115, 9981 –10080; e) H. Lu, P. Zhang, Chem. Soc.
Rev. 2011, 40, 1899-1909; f) M. P. Doyle, R. Duffy, M. Ratnikov, L. Zhou,
Chem. Rev. 2010, 110, 704−724.
[17] a) F. Milocco, S. Demeshko, F. Meyer, E. Otten, Dalton Trans., 2018, 47,
8817–8823; b) R. Travieso-Puente, J. O. P. Broekman, M.-C. Chang, S.
Demeshko, F. Meyer, E. Otten, J. Am. Chem. Soc. 2016, 138,
5503−5506; c) W. D. Morris, P. T. Wolczanski, J. Sutter, K. Meyer, T. R.
Cundari, E. B. Lobkovsky, Inorg. Chem. 2014, 53, 7467-7484; d) L.
Zhang, L. Xiang, Y. Yu, L. Deng, Inorg. Chem. 2013, 52, 5906-5913; e)
C: A. Nijhuis, E. Jellema, T. J. J. Sciarone, A. Meetsma, P. H. M.
Budzelaar, B. Hessen, Eur. J. Inorg. Chem. 2005, 2089-2099; f) A.
Panda, M. Stender, R. J. Wright, M. M. Olmstead, P. Klavins, P. P. Power,
Inorg. Chem. 2002, 41, 3909-3916; g) B. Vendemiati, G. Prini, A.
[3]
[4]
Y.-R. Luo, Comprehensive Handbook of Chemical Bond Ener-gies, CRC
Press, Boca Ratón, 2007.
See reference 2 and for recent examples: a) A. E. Shiely, L.-A Clarke, C.
J. Flynn, A. M. Buckley, A. Ford, S. E. Lawrence, A. R. Maguire, Eur. J.
Org. Chem., 2018, 19, 2277-2289; b) Y. Wang, X. Wen, X. Cui, X. P.
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
Meetsma, B. Hessen, J. H. Teuben, O. Traverso, Eur. J. Inorg. Chem.
2001, 707-711; h) J. R. Hagadorn, J. Arnold, Inorg. Chem. 1997, 36, 132-
133.
[18] T. Janes, J. M. Rawsonb, D. Song, Dalton Trans., 2013, 42, 10640-
10648.
[19] a) DeAngelis, A.; Panish, R.; Fox, J.M. Acc. Chem. Res. 2016, 49,
115−127 and references cited in the review; b) D. F. Taber, M. J.
Hennessy, J. P. Louey, J.Org. Chem. 1992, 57,436-441; c) M. Lankelma,
A. M.Olivares, B. de Bruin, Chem. Eur. J., 2019, 25, 5658 –5663.
[20] D. M. Seo, T. Afroz, J. L. Allen, P. D. Boyle, P. C. Trulove, H. C. De Long,
W. A. Henderson, J. Phys. Chem. C 2014, 118, 25884−25889.
[21] M. M. Díaz-Requejo, P. J. Pérez, Chem. Rev. 2008, 108, 3379-3394, and
references cited in the review.
[22] M. Gómez-Gallego, M. A. Sierra, Chem. Rev. 2011, 111, 4857–4963.
[23] G. A. Sulikowski, S. Lee, Tetrahedron Lett. 1999, 40, 8035-8038.
[24] C. D.-T. Nielsen, J. Burés, Chem. Sci., 2019, 10, 348–353.
This article is protected by copyright. All rights reserved.

10.1002/anie.201905986
Angewandte Chemie International Edition
RESEARCH ARTICLE
RESEARCH ARTICLE
Alberto Hernán-Gómez,*
Mònica Rodríguez, Teodor
Parella, Miquel Costas*
Iron-carbene gets access to strong Csp3-H bonds: Pairing a strong lewis acidic
iron catalyst with pre-activation of diazo reagents using the lithium salt LiAl(ORF)4
[ORF= (OC(CF3)3] renders electrophilic iron-carbenes capable to insert in
non-activated aliphatic C-H bonds.
Page No. – Page No.
Electrophilic iron catalyst paired with
a
lithium cation enables selective
functionalization of non-activated
aliphatic C-H bonds via
metallocarbene intermediates
[a]
[b]
A. Hernán-Gómez, M. Rodríguez, M. Costas
Departament de Química
Universitat de Girona, Institut de Química Computacional i
(IQCC)
C/ M. Aurèlia Capmany 69, 17003 Girona, Catalonia (Spai
E-mail: miquel.costas@udg.edu
T. Parella
This article is protected by copyright. All rights reserved.
Departament de Química