Iron(III) triflimide as a catalytic substitute for gold(I) in hydroaddition reactions to unsaturated carbon-carbon bonds

Article abstract of DOI:10.1002/chem.201300386

In this work it is shown that iron(III) and gold(I) triflimide efficiently catalyze the hydroaddition of a wide array of nucleophiles including water, alcohols, thiols, amines, alkynes, and alkenes to multiple C-C bonds. The study of the catalytic activity and selectivity of iron(III), gold(I), and Bronsted triflimides has unveiled that iron(III) triflimide [Fe(NTf ₂)₃] is a robust catalyst under heating conditions, whereas gold(I) triflimide, even stabilized by PPh₃, readily decomposes at 80 °C and releases triflimidic acid (HNTf₂) that can catalyze the corresponding reaction, as shown by in situ ¹⁹F, ¹⁵N, and ³¹P NMR spectroscopy. The results presented here demonstrate that each of the two catalyst types has weaknesses and strengths and complement each other. Iron(III) triflimide can act as a substitute of gold(I) triflimide as a catalyst for hydroaddition reactions to unsaturated carbon-carbon bonds. Lewis and Bronsted catalysis: It is shown that iron(III) catalyses as efficiently as gold(I) the hydroaddition of a wide array of nucleophiles, including water, alcohols, thiols, amines, alkynes, and alkenes, to multiple C-C bonds (see scheme). Copyright

Full text of DOI:10.1002/chem.201300386

FULL PAPER
DOI: 10.1002/chem.201300386
Iron
ACHTUNGTRENNUNG
Jose R. Cabrero-Antonino, Antonio Leyva-Pꢀrez,* and Avelino Corma*[a]
Abstract: In this work it is shown that
iron(III) and gold(I) triflimide effi-
ciently catalyze the hydroaddition of a
wide array of nucleophiles including
water, alcohols, thiols, amines, alkynes,
robust catalyst under heating condi-
tions, whereas gold(I) triflimide, even
stabilized by PPh3, readily decomposes
at 808C and releases triflimidic acid
(HNTf2) that can catalyze the corre-
sponding reaction, as shown by in situ
19F, 15N, and 31P NMR spectroscopy.
The results presented here demonstrate
that each of the two catalyst types has
weaknesses and strengths and comple-
ACHTUNGTRENNUNG
À
and alkenes to multiple C C bonds.
The study of the catalytic activity and
ment each other. IronACTHNGUTERNNU(G III) triflimide
can act as a substitute of gold(I) trifl-
imide as a catalyst for hydroaddition
reactions to unsaturated carbon–
carbon bonds.
Keywords: Brønsted catalysis
gold hydroaddition reactions
iron · Lewis catalysis
·
·
selectivity of iron
Brønsted triflimides has unveiled that
iron(III) triflimide [Fe(NTf2)3] is a
ACHTUNGTRENNUNG(III), gold(I), and
·
A
U
Introduction
Results and Discussion
Modern organic synthesis based on catalysis by transition
metals has increased its possibilities with the introduction of
gold for making new carbon–carbon and carbon–heteroatom
bonds.[1–13] Nonetheless, there is much incentive to substitute
expensive and toxic heavy late transition metals by cheaper,
widely available, and environmentally-friendly first-row
transition metals, such as iron.[14–28] The aim of our work
here is to study iron, gold, and triflimidic acid, representa-
tive examples of first-row and heavy late transition metals
and Brønsted acids, respectively, as catalysts for hydroaddi-
Stability of triflimide salts: The traditional division into
Brønsted and Lewis catalysis is difficult to be separated
when in situ generated protons can contribute to the catalyt-
ic activity of some metals.[31,32] Triflimide metal salts [Mn+
AHCTUNTGRENNG(UN NTf2)n], particularly those with gold, have gained much at-
tention in the last years[15,16,21,33–36] due to their high Lewis
acidity and easy handling relative to the corresponding trifl-
imidic acid (HNTf2). Despite the fact that these salts are
readily hydrolizable, the possible role of acid protons on the
final catalytic activity is not always tested.[37–43] This is not
surprising since unveiling the role of protons for metal-cata-
lyzed reactions (and particularly for gold) is not trivial, be-
cause the claimed water tolerance of gold catalysts has led
to a “wet” chemistry, early recognized by Hashmi as a possi-
ble source of catalytic protons.[31,32] Blank experiments with
the corresponding acid and addition of the non-coordinating
base di(tert-butyl)pyridine are the two ways commonly em-
ployed to detect a possible Brønsted catalysis.[15,44] However,
recent studies have shown that these methods sometimes
fail when other factors come into play as we will discuss
below.[37–43] For instance, it has been shown that the hydroa-
mination[43] of nonactivated alkenes proceeds with catalytic
amounts of triflic acid (HOTf) below 5 mol%, since higher
amounts (20 mol%) only trigger the decomposition of the
substrate.[38,40,41] Related hydroalkoxylation reactions only
proceed at temperatures below 508C when they are cata-
lyzed by triflic acid since higher temperatures decompose
the substrate. In contrast, the corresponding gold(I) triflate
À
tion reactions to p-carbophilic unsaturated C C bonds (al-
kenes and alkynes) of carbon, oxygen, nitrogen, and sulfur
nucleophiles, a set of complete atom-economical reactions
with growing potential in organic synthesis.[3,8,29,30] The study
includes reactivity tests, isotopic labeling, in situ NMR and
EPR experiments, and cyclic voltammetry, which have
helped to shed light on the exact nature of the catalytic
process and active sites involved.
[a] J. R. Cabrero-Antonino, Dr. A. Leyva-Pꢀrez, Prof. Dr. A. Corma
Instituto de Tecnologꢁa Quꢁmica
Consejo Superior de Investigaciones Cientꢁficas-Universidad
Politꢀcnica de Valencia
Avda. de los Naranjos s/n, 46022 Valencia (Spain)
Fax : (+34)963879444
(PPh3)] works nicely at 858C.[40] On the other
[AuACTHNUGRTENNUG(OTf)ACHTUGNTRENNUGN
hand, the addition of di(tert-butyl)pyridine can generate
rather than neutralize catalytic protons under specific reac-
tion conditions, which is an opposite effect to the one ex-
pected. All these examples illustrate that Brønsted catalysis
E-mail: anleyva@itq.upv.es
acorma@itq.upv.es
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201300386.
Chem. Eur. J. 2013, 19, 8627 – 8633
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8627

can be hidden by different factors that are only assessed
through specifically designed experiments. In our case, we
will show how the potential generation of protons from the
triflimide salts can be systematically monitored by in situ
19F, 15N, and 31P NMR spectroscopic experiments to deter-
mine the exact nature of the catalysis.
Hydroaddition reactions to alkynes: The results for the iron-
AHCTUNGERTG(NNUN III), gold(I), or Brønsted triflimide-catalyzed hydroaddition
of oxygen (reaction A), sulfur (reactions B and C), and
carbon (reaction D) nucleophiles to alkynes are shown in
Figure 3. Additional reactions and catalysts together with a
Prior to that, the stability of these FeIII and AuI triflimide
species versus different nucleophiles was examined by
19F NMR spectroscopy. Figure 1 shows that the addition of
Figure 3. Catalytic study of ironACTHNUGRTENUNG(III), gold(I), and Brønsted triflimides for
various hydroaddition reactions to phenylacetylenes. Gold(I) triflimide
Figure 1. 19F NMR spectra in [D8]1,4-dioxane of the iron
(NTf2)3] after the addition of different nucleophiles at room temperature.
ACHTUNGTREN(NUNG III) salt [Fe-
1
À
ACHTUNGTRENNUNG
was used as a [Au
H2 =H2O (3 equiv), 20 mol% of [Fe
G
E
(NTf2)3] and 60 mol% of HNTf2
were used. Reaction B: R =H 9 (2 equiv), Nu H2 =benzene-1,2-dithiol
10. Reaction C: R =Et 4 (2 equiv), Nu H2 =benzene-1,2-dithiol 10. Re-
action D: R =H 9, Nu H2 =mesitylene 20 (5 equiv), dichloromethane as
the solvent and 30 mol% of HNTf2. GC conversions and yields were cal-
culated by using dodecane as an external standard.
1
1
H2O, PhNH2, PhOH, and PhSH shifts the 19F NMR spectro-
scopic signal of [Fe(NTf2)3]. However, the broad singlet
AHCTUNGTRENNUNG
found in most cases suggests that no hydrolysis occurs but
only the formation of the corresponding Fe nucleophile ad-
ducts.
complete set of numerical data can be seen in the Support-
ing Information (Tables S1–S7). The conversion (X) and the
regioisomer distribution are indicated for each reaction and
catalyst.
In the case of [AuACTHNUTRGENNUG(NTf2)CAHTUNGTREN(NUGN Ph3P)], Figure 2 shows that the
addition of an excess of water does not produce any change
to the catalyst signal after heating at 808C over 3 h but that
the stability of gold towards hydrolysis readily disappears
when a better ligand, such as an alkyne is added.
The results show that the hydration[16] of 1-phenyl-1-pro-
pyne 1 (reaction A) under heating conditions proceeds
cleanly to the Markovnikov product (M) 2 with [Fe
ACHTUNGTRENNUNG(NTf2)3]
and HNTf2 as catalysts, whereas [Au(NTf2)(PPh3)] gives a
A
ACHTUNGTRENNUNG
mixture of Markovnikov (M) 2 and anti-Markovnikov (A-
M) 3 products. In situ 19F NMR spectroscopy reveals that
the ironACHTNUTRGNE(UNG III) triflimide does not hydrolyze to the Brønsted
acid, whereas the gold(I) triflimides readily do, as can be
seen in Figures 4 and 5 after comparing the original values
of the catalysts at d=À73 and À76 ppm with the value of
triflimidic acid at d=À79 ppm under the same reaction con-
ditions. Differences in selectivity can be explained by con-
sidering that, whereas iron smoothly forms the intermediate
adduct that catalyzes the regioselective reaction, gold cata-
lyzes the unselective hydration and then decomposes.[9,16,45]
In situ 31P NMR spectroscopy[46] (see Figure S1 in the Sup-
porting Information) confirmed the decomposition of gold
with formation of a black precipitate at the end of the reac-
tion. This precipitate was isolated, weighted, and analyzed
by inductively-coupled plasma mass spectrometry (ICP-MS)
to establish the gold mass balance.
Figure 2. 19F NMR spectra in [D8]1,4-dioxane of the gold(I) complex [Au-
A
ACHTUNGTRENNUNG(PPh3)] after the addition of an excess of water at two different
8628
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 8627 – 8633

Hydroaddition Reactions to Unsaturated Carbon–Carbon Bonds
FULL PAPER
To further confirm these observations, isotopically-labeled
15N-silver triflimide 8 was synthesized[47] by using 15N-benzyl-
amine 7b as the starting material, as depicted in Scheme 1,
and employed for the synthesis of 15N-labeled iron
ACHTUNGTRENNUNG(III),
gold(I), and Brønsted triflimides, which were used as cata-
lysts while following the reaction by in situ 15N NMR spec-
troscopy. The results in Figures 6 and 7 confirmed that the
Figure 4. In situ 19F NMR spectra for the hydration of 1-phenyl-1-butyne
4 with H2O (3 equiv) by using [Fe
tivity 5/6.
ACHTUNGTERN(NUNG NTf2)3] as the catalyst. Rs: regioselec-
Figure 6. In situ 15N NMR spectra for the hydration of 1-phenyl-1-butyne
4 with H2O (3 equiv) by using [FeACHTUNGTRNEUNG
(15NTf2)3] as the catalyst.
active species for ironACTHNUTRGNE(NUG III) triflimide is not triflimidic acid
(compare the value of the catalyst during the reaction at d=
148 ppm with that observed for H15NTf2 at d=139 ppm, see
also Figures S2–S5), whereas the gold(I) triflimide complex
is readily hydrolyzed under the reaction conditions.
It is striking that the stable [AuACTHNUGTRENNGU(NTf2)ACHTUNTGREN(NNGU PPh3)] gold(I) com-
plex reduces so easily to gold(0) in the presence of a steri-
cally demanding alkyne, such as 1-phenyl-1-butyne 4 and a
weak-bounded nucleophile, such as water, even under heat-
ing conditions. Thus, it is reasonable to expect that many un-
Figure 5. In situ 19F NMR spectra for the hydration of 1-phenyl-1-butyne
4 with H2O (3 eqꢃiv) by using [Au
gioselectivity 5/6.
ACHTUNGTERNNUG(NTf2)ACHTUNGTNER(NUGN PPh3)] as the catalyst. Rs: re-
À
saturated C C bonds and nucleophiles will promote the de-
composition of the gold(I)
complex.[48,49] In the case of the
dihydrothiolation[50] of a simple
terminal alkyne, such as phe-
nylacetylene 9 with benzene-
1,2-dithiol 10, as a nucleophile
(Figure 3,
reaction B),
the
three catalysts showed excel-
lent conversion (>99%), but
only ironACTHNUGRTENUNG(III) triflimide gave
excellent yields to the Markov-
nikov product (90% of 11),
whereas gold(I) gave a lower
Scheme 1. Synthesis of isotopically labeled [Ag15NTf2] 8 by using 15N-benzylamine 7b as the starting material.
Chem. Eur. J. 2013, 19, 8627 – 8633
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8629

A. Leyva-Pꢀrez, A. Corma and J. R. Cabrero-Antonino
Table 1. Dihydrothiolation of alkynes with [FeACTHUNRTGNEUNG(NTf2)3] as a catalyst.
Run[a]
1
Alkyne
Main product
Yield [%][b]
71 (65)
2
98 (76)
3
4
88 (74)
92 (83)
[a] Reaction conditions: alkyne (0.50 mmol), benzene-1,2-dithiol 10
(46.0 mL, 0.25 mmol), FeCl3 (4.0 mg, 0.025 mmol), [AgNTf2] (30.0 mg,
0.075 mmol), anhydrous 1,4-dioxane (0.5 mL) at 808C for 24 h. [b] GC
yield shown in parentheses with dodecane as an external standard.
Figure 7. In situ 15N NMR spectra for the hydration of 1-phenyl-1-butyne
4 with H2O (3 equiv) by using [Au(NTf2)(PPh3)] as the catalyst.
N
ACHTUNGTRENNUNG
selectivity (79% of 11) and the Brønsted triflimide gave the
anti-Markovnikov product (68% of 12).
with ironACTHNGUTRENNU(G III) triflimide with respect to gold comes from the
higher amount of HNTf2 generated.
All these results point to a general decomposition of the
In situ 19F NMR spectra show that the iron
(III) triflimide
does not decompose or hydrolyze under the reaction condi-
tions (see Figure S6 in the Supporting Information), whereas
the gold(I) triflimide does (Figure S7), since a new signal at
d=À79 ppm appears after 15 min of reaction and is very
similar to that observed for the Brønsted acid under the
same reaction conditions. Working with another alkyne,
such as 1-phenyl-1-butyne 4 (Figure 3, reaction C), it was
found again that ironACHTUNRGTNE(NUG III) triflimide is a better catalyst
(86% of Markovnikov product 13) than gold(I) (79% of 13)
or the Brønsted triflimide (45% of 14).
With these catalytic results in hand, we expanded the
ironACHTUNGTRENNUNG(III)-catalyzed dihydrothiolation to other alkynes by
using benzene-1,2-dithiol 10 as the nucleophile and the re-
sults in Table 1 show that aromatic (Table 1, entries 1–3)
and aliphatic alkynes (entry 4) are converted to dithioacetals
[AuACHTUNGRTENNU(G NTf2)ACHTUNGTREN(NGNU PPh3)] complex in the presence of C–C multiple
bonds under heating conditions. However, it must be noted
that gold(I) is also active at room temperature for intermo-
lecular hydroadditions[46,52] to alkynes, whereas iron
ACHTUNGTRENNUNG(III) tri-
flimide is inactive (see Tables S5 and S6 in the Supporting
Information). As could be expected, in-situ NMR spectro-
scopic studies revealed that [Au
at lower temperatures,[46] but a significant degree of decom-
position with time can be expected. In addition, iron(III) tri-
ACHUTNGTRENGUN(NTf2)ACHTUGNTREN(NUGN PPh3)] is more stable
ACHTUNGTRENNUNG
flimide is also active for intramolecular hydroadditions, such
as, for instance, the hydroalkoxylation of internal alkynes,
which is the most used gold-catalyzed transformation in
total synthesis to date (see Table S7 in the Supporting Infor-
mation).[1,3,4,53]
with 10 mol% of [Fe
Markovnikov selectivity.
G
Hydroaddition reactions to alkenes: Complementarily to the
study of alkynes, the results for the hydroaddition of sulfur
(reaction A), carbon (reactions B and C), nitrogen (reac-
tion D), and oxygen (reaction E) nucleophiles to alkenes by
using catalytic amounts of the three triflimide species are
shown in Figure 8. For complete numerical information for
these and more catalysts and additional reactions tested, see
Tables S8–S17 in the Supporting Information.
When a carbon nucleophile, such as mesitylene 20, was
used for the hydroarylation[51] of phenylacetylene
(Figure 3, reaction D) the results showed that the Brønsted
triflimide gave the best results (>99% conversion and 91%
yield of 21) and in situ 19F NMR spectroscopy reveals that
both iron and gold triflimide hydrolyze under the reaction
conditions (see Figures S8 and S9 in the Supporting Infor-
mation), which explains the similar catalytic performance
for the three triflimides. 31P NMR spectroscopy confirmed
the decomposition of the gold(I) complex (see Figure S10 in
the Supporting Information). The better result obtained
The results for the hydrothiolation[15] of 4-chlorostyrene
22 with thiophenol 23 (Figure 8, reaction A) indicate that
the ironACHTUNGTRNEUNG
(III) triflimide is the best catalyst, and in situ 19F and
15N NMR spectroscopies (see Figures S11 and S12 in the
Supporting Information) confirmed that it does not hydro-
8630
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 8627 – 8633

Hydroaddition Reactions to Unsaturated Carbon–Carbon Bonds
FULL PAPER
thinylation[23] of the bicyclic alkene norbornene 24 (Figure 8,
reaction C) since ironACTHNUGRTENUNG(III) and Brønsted triflimide gave
again the best results, whereas gold(I) was significantly less
active. The in situ 19F NMR spectroscopic experiments (see
Figure S18 in the Supporting Information) with ironACHTUNGTRENNUNG(III) tri-
flimide confirms that this species does not hydrolyze during
the reaction. Hydroadditions[41] of nitrogen (tosylamine 25,
Figure 8, reaction D) and oxygen (p-methoxybenzoic acid
26, reaction E) nucleophiles to norbornene 24 were also
studied and Brønsted acids were the best catalysts (86 and
82% yield for HOTf, and 70 and 78% yield for HNTf2, re-
spectively). These observations are in good agreement with
the results reported by other authors.[41]
Some conclusions about the stability and catalytic activity
of the iron
ACHTUNGTRENNUNG
from the results presented above. [Au
N
ACHTUNGTRENNUNG
decomposes in the presence of alkynes and alkenes under
heating conditions, which contrasts with the stability towards
water shown in Figure 2. We can say that decomposition of
Figure 8. Catalytic study of iron
various hydroaddition reactions to alkenes. Gold(I) triflimide was used as
the [AuACHTUNGTRENNUNG(NTf2)ACHTUNGTRENNUNG(PPh3)] complex. Reaction A: alkene=4-chlorostyrene 22,
ACHTUNGTRENUN(NG III), gold(I), and Brønsted triflimides for
(PPh3)] occurs mainly by gold reduction[48,49] (al-
[AuACTHNUTRGENN(UG NTf2)ACHTUNGTRENNUGN
À
Nu H=thiophenol 23 (1 equiv). Reaction B: alkene=4-chlorostyrene
À
À
though sequestration by high-affinity ligands, such as thiols
can also occur) and, in consequence, gold is significantly less
active than Fe3+ and H+ as a catalyst under these reaction
conditions. This behavior also occurs with alkenes, which in-
teract better with harder Lewis acids, such as FeIII. The en-
22, Nu H=4-chlorostyrene 22. Reaction C: alkene=norbornene, Nu
H=phenylacetylene 9 (4 equiv), 1,2-dichloroethane as the solvent and
À
160 h reaction time. Reaction D: alkene=norbornene 24, Nu H=tosyl-
amine 25 (1 equiv), 1,2-dichloroethane as the solvent and 6 h reaction
time. Reaction E: alkene=norbornene 24, Nu H=p-methoxybenzoic
acid 26 (1 equiv). GC conversions and yields were calculated by using do-
À
decane as an external standard.
hanced Lewis acidity of [FeACHTNUGTRENUNG(NTf2)3] towards alkynes and al-
kenes comes from the combination of a hard Lewis center
with a soft low-coordinating anion.[15,16,21,35] According to
Pearsonꢄs (hardness–softness) rules, hard cations bounded to
soft anions become softer and it is not surprising that iron
triflimides show an improved ability to activate (C–C) unsa-
turated double bonds.[35] Incidentally, ferric triflimide [Fe-
lyze under the reaction conditions. Gold(I) gave poorer re-
sults since it hydrolyzes under reaction conditions as con-
firmed by 19F, 31P, and 15N NMR spectroscopic studies (see
Figures S13–15 in the Supporting Information). The poorer
activity of gold(I) in comparison with iron
A
(NTf2)3] presents a stabilized 3d5 high-spin (hs) semifilled
ACHTUNGTRENNUNG
bed to sulfur poisoning and the higher affinity of Au for S
configuration as confirmed[16] by UV spectroscopy and EPR
(see Figures S19–20 in the Supporting Information), which
resembles in some way the 5d10 stabilization of Au+ and
Hg2+[54] and produces the softening of the Fe3+ cation (ionic
radius: 64.5 pm for hs vs. 55.0 pm for ls). We expect that the
increase in the ionic radius of iron would be reflected in the
stabilization of the LUMO, which ultimately dictates the
À1
À1
À
À
(Au S=418 kJmol ) relative to Fe (Fe S=322 kJmol )
would explain the catalytic differences observed with al-
kynes (see dihydrothiolation) and alkenes. In other words,
besides gold(I) reduction there is a second way of gold(I)
deactivation by sulfur poisoning, as evidenced by the small
number of efficient gold-catalyzed transformations involving
sulfur nucleophiles.[3,5,8,49] Interestingly, we have shown that
catalytic activity of the Lewis acids in hydroaddition reac-
[35]
À
iron
G
tions to unsaturated C C bonds. Cyclic voltammetry ex-
to alkynes and alkenes.[15]
periments were performed for iron
(III) triflimide, triflate,
If no additional nucleophiles are added to the reaction
mixture, the alkene itself can act as a nucleophile and the
head-to-tail dimerization[21] (or hydrovinylation) of 4-chlor-
ostyrene 22 (Figure 8, reaction B) proceeds very well in the
and chloride (see Figure S21 in the Supporting Information)
and the reduction potential values for [Fe3+/Fe2+] in Table 2
show a progressive decrease from chloride to triflimide, con-
firming the LUMO stabilization[55] of iron
ACHTUNGTRENNUNG
presence of either [Fe
G
with respect to ironACTHNGUTRENNU(G III) triflate or ironACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
spectroscopy (see Figures S16 and S17 in the Supporting In-
formation) indicates that iron does not hydrolyze to HNTf2
under the reaction conditions, so the dimerization of alkenes
can also be classified as a hard Lewis-catalyzed reaction,
since Fe3+ and H+ perform better than Au+. It is not sur-
prising that harder acidic centers are needed to activate al-
kenes since they are considered harder electrophiles than al-
kynes. This trend was confirmed after studying the hydroe-
Table 2. [Fe3+/Fe2+] reduction potential values for three iron
ACTHNUGRTENUNG(III) salts.
Iron
G
E0 [Fe3+/Fe2+
]
FeCl3
0.52 V
0.50 V
0.42 V
[Fe
[Fe
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Chem. Eur. J. 2013, 19, 8627 – 8633
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8631

A. Leyva-Pꢀrez, A. Corma and J. R. Cabrero-Antonino
Typical reaction procedure for the dihydrothiolation of phenylacetylenes
(see Tables S2–S3 in the Supporting Information): The corresponding cat-
alyst was placed in a 2 mL vial and a rubber septum was fitted. Dry 1,4-
dioxane (0.5 mL), phenylacetylene (0.5 mmol), and benzene-1,2-dithiol
10 (46 mL, 0.25 mmol) were added and the mixture was placed in a pre-
heated oil bath at room temperature and magnetically stirred for 24 h.
After cooling, n-hexane (1 mL) was added, the liquid was passed through
a microfilter syringe, and the filtrates were analyzed by GC analysis after
addition of dodecane (11.0 mL, 0.048 mmol) as the external standard.
Synthesis of 15N-benzyl sulfonamide (7c) (Scheme 1): DIPEA (2.46 mL,
14.18 mmol) was added to a magnetically stirred solution of 15N-benzyl-
amine 7b (776.5 mL, 7.1 mmol) in dry DCM (70 mL) under a nitrogen at-
mosphere. The reaction mixture was cooled down À788C, and then triflic
anhydride 7a (5.0 g, 17.73 mmol) was added dropwise. The reaction mix-
ture was slowly warmed up to room temperature over 1 h and stirred for
1 h. Then, aqueous HCl (3%) was added. The aqueous phase was ex-
tracted with dichloromethane (DCM) and the combined organic layers
were dried over MgSO4. Filtration and evaporation of the solvent left a
crude mixture, which was refluxed in pentane. Before cooling down to
room temperature, the pentane phases were collected by repeating this
operation several times. Evaporation of the pentane fractions left N-
benzyl sulfonamide 7c as a pale-brown solid (2.17 g, 83%). 1H NMR
(300 MHz, CDCl3): d=7.52–7.46 (2H, m), 7.43–7.37 (3H, m), 5.09 ppm
(2H, s); 13C NMR (300 MHz, CDCl3): d=131.7 (C), 129.9 (CH), 129.8
(2ꢅCH), 128.9 (2ꢅCH), 118.8 (CF3, q, J1CÀF =324.9), 56.6 ppm (CH2).
Conclusion
Iron
lar hydroaddition to C–C multiple bonds of carbon, oxygen,
sulfur, and nitrogen nucleophiles. [Au(PPh3)(NTf2)] readily
ACHTUNGTRENNUNG(III) triflimide is an active catalyst for the intermolecu-
G
ACHTUNGTRENNUNG
decomposes in the presence of alkynes and alkenes under
heating conditions and the decomposition occurs mainly by
gold reduction or sequestration by high-affinity ligands such
as thiols. As a consequence, iron is a more active catalyst
than gold for these reactions at temperatures ~808C and the
use of ironACHTUNGTRENNUNG(III) triflimide, until now anecdotic in organic
catalysis, could be a good alternative to the use of late
heavy transition metals for hydroaddition reactions.
Experimental Section
General: Reagents and solvents were obtained from commercial sources
and were used without further purification unless otherwise indicated.
All the products obtained were characterized by GCMS analysis, H- and
1
13C NMR spectroscopy, and DEPT. When available, the characterization
given in the literature was used for comparison. Gas chromatographic
analyses were performed in an instrument equipped with a 25 m capillary
column of 5% phenylmethylsilicone. GCMS analyses were performed on
a spectrometer equipped with the same column as the GC and operated
under the same conditions. 1H-, 13C-, DEPT, 19F-, 31P-, and 15N NMR
spectroscopic measurements were recorded in a 300 MHz instrument by
using CDCl3 as the solvent containing TMS as an internal standard. For
the 19F NMR spectroscopic experiments, fluorobenzene was used as the
internal standard. Absorption spectra were recorded on an UV/Vis spec-
trophotometer (UV0811M209, Varian) and EPR experiments were car-
ried out with Bruker EMX/X equipment. IR spectra of the compounds
were recorded on a Jasko 460 plus spectrophotometer by impregnating
the windows with a dichloromethane solution of the compound and leav-
ing it to evaporate before analysis. IR peaks are defined as: very intense
(vi), intense (i), medium (m), low (l), and broad (br). Cyclic voltammetry
measurements were carried out by using the conventional three-electrode
setup connected to an Amel potentiostat (model 7050) that was control-
led by software allowing data storage and management. Platinum wire,
gold wire, and Ag/AgCl electrode (0.1m KCl standard solution) were
used as counter, working, and reference electrodes, respectively. The
electrodes were immersed into a N2 purged 0.1m tetrabutylammonium
perchlorate electrolyte solution in acetonitrile. Measurements were car-
ried out at a scan rate of 20 mVsÀ1 in the range À2.0/+2.0 V.
Synthesis of 15N-silver(I) triflimide (8) (Scheme 1): N-Benzyl sulfonamide
7c (1.19 g, 3.20 mmol) was dissolved in ethanol (16 mL) and stirred over
8 h at room temperature. The volatiles were evaporated from the reac-
tion mixture under reduced pressure at 608C. The oil residue was dis-
solved in dry toluene (25 mL). Then, silver oxide (0.371 g, 1.6 mmol) was
added and the light-protected reaction mixture was heated under reflux
for 3 h, after which complete dissolution of the solid was observed. The
reaction mixture was cooled down, filtered over Celite and concentrated
to 1/3 of the volume. Finally, the product [Ag15NTf2] 8 was precipitated
with pentane as a yellow hygroscopic solid (0.87 g, 70%). 19F NMR
(300 MHz, [D8]1,4-dioxane): d=À77.2 ppm (s, CF3); 15N NMR
(300 MHz, [D8]1,4-dioxane): d=141.7 ppm (s, 15NTf2).
À
Typical reaction procedure for in situ 15N NMR spectroscopic experi-
ments for the hydration of 1-phenyl-1-butyne (4) catalyzed by [Fe-
AHCTUNGTRENNUNG
(15NTf2)3] (Figure 6). FeCl3 (97%, Aldrich, 8.0mg, 0.05mmol) and
[Ag15NTf2] (60.0mg, 0.15mmol) were placed in a 2mL vial. A rubber
septum was fitted and dry [D8]1,4-dioxane (1.0mL) was added. The mix-
ture was magnetically stirred at room temperature for 30min. Then, 1-
phenyl-1-butyne 4 (72.0 mL, 0.5mmol) and water (28 mL, 1.5mmol) were
added and the mixture was placed in preheated oil bath at 1208C and
magnetically stirred for 20h. The reaction mixture was periodically ana-
lyzed by 15N NMR spectroscopy at several points during the reaction.
One aliquot (10 mL) was diluted in n-hexane (1mL) and passed through
a microfilter syringe. Then, dodecane (11 mL, 0.048mmol) was added as
an external standard and the filtrates were analyzed by GC analysis.
Typical procedure for in situ 19F NMR spectra (Figure 1): FeCl3 (97%,
Aldrich, 8.0 mg, 0.05 mmol) and [AgNTf2] (97%, Aldrich, 60.0 mg,
0.15 mmol) were placed in a 2 mL vial. A rubber septum was fitted and
dry [D8]1,4-dioxane (1.0 mL) was added. The mixture was magnetically
stirred at room temperature for 30 min observing the precipitation of
AgCl. Then, the liquid was passed through a microfilter syringe, and the
filtrates were placed in a 2 mL vial. Nucleophile (0.2 mmol; water, ani-
line, phenol, and thiophenol) was added in each case. The resultant mix-
ture was magnetically stirred at room temperature over 5 min and ana-
lyzed by 19F NMR spectroscopy by using fluorobenzene as the internal
standard.
Typical reaction procedure for dihydrothiolation reactions (Table 1):
FeCl3 (97%, Aldrich, 8.0 mg, 10 mol%) and [AgNTf2] (97%, Aldrich,
60.0 mg, 30 mol%) were placed in a 10 mL round-bottomed flask. Dry
1,4-dioxane (1.5 mL) was added and the mixture was magnetically stirred
at room temperature for 30 min. Then, alkyne (1.0 mmol) and benzene-
1,2-dithiol 10 (92 mL, 0.5 mmol) were added and the mixture was placed
in a pre-heated oil bath at 808C and magnetically stirred for 24 h. After
cooling, n-hexane (10 mL) was added, observing the precipitation of the
catalyst. The liquid was passed through a microfilter syringe and one ali-
quot of the filtrates was analyzed by GC analysis after addition of dodec-
ane (11.0 mL, 0.048 mmol) as the external standard. Finally, the liquid
was purified by preparative TLC by using the corresponding eluent.
Typical reaction procedure for hydration of 1-phenyl-1-propyne (1) (see
Table S1 in the Supporting Information): The corresponding catalyst was
placed in a 2 mL vial and a rubber septum was fitted. 1,4-dioxane
(0.5 mL), 1-phenyl-1-propyne 1 (31.3 mL, 0.25 mmol), and water (14 mL,
0.75 mmol) were added and the mixture was placed in a pre-heated oil
bath at the corresponding temperature and magnetically stirred for 20 h.
After cooling, n-hexane (1 mL) was added, the liquid was passed through
a microfilter syringe and the filtrates were analyzed by GC analysis after
addition of dodecane (11.0 mL, 0.048 mmol) as the external standard.
Typical procedure for cyclic voltammetry (Table 2 and Figure S21 in the
Supporting Information): MeCN solutions of FeCl3 (0.5 mm) were pre-
pared as follows: FeCl3 (>99.99%, Aldrich, 4.0 mg, 0.025 mmol) was di-
luted in MeCN (5 mL). Then, 500 mL of this solution was diluted in
MeCN (5 mL) and purged with N2 over 15 min. Cyclic voltammetry
8632
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 8627 – 8633

Hydroaddition Reactions to Unsaturated Carbon–Carbon Bonds
FULL PAPER
measurements were carried out by using the equipment described in the
General section.
[22] S. Zhou, K. Junge, D. Addis, S. Das, M. Beller, Angew. Chem. 2009,
121, 9671; Angew. Chem. Int. Ed. 2009, 48, 9507.
[23] K. Kohno, K. Nakagawa, T. Yahagi, J.-C. Choi, H. Yasuda, T. Saka-
kura, J. Am. Chem. Soc. 2009, 131, 2784.
[24] A. Correa, O. Garcia Mancheno, C. Bolm, Chem. Soc. Rev. 2008, 37,
1108.
[25] J. Michaux, V. Terrasson, S. Marque, J. Wehbe, D. Prim, J.-M. Cam-
pagne, Eur. J. Org. Chem. 2007, 2601.
[26] C. Bolm, J. Legros, J. Le Paih, L. Zani, Chem. Rev. 2004, 104, 6217.
[27] A. Fꢈrstner, A. Leitner, M. Mendez, H. Krause, J. Am. Chem. Soc.
2002, 124, 13856.
[28] J. Kischel, I. Jovel, K. Mertins, A. Zapf, M. Beller, Org. Lett. 2006,
8, 19.
[29] N. T. Patil, R. D. Kavthe, V. S. Shinde, Tetrahedron 2012, 68, 8079.
[30] M. Beller, J. Seayad, A. Tillack, H. Jiao, Angew. Chem. 2004, 116,
3448; Angew. Chem. Int. Ed. 2004, 43, 3368.
Typical procedure for cyclic voltammetry experiments for [Fe
(Table 2 and Figure S21 in the Supporting Information): MeCN solutions
of [Fe(NTf2)3] (0.1 mm) were prepared as follows: FeCl3 (>99.99%, Al-
ACHTUNGTRNE(NUNG NTf2)3]
ACHTUNGTRENNUNG
drich, 4.0 mg, 0.025 mmol,) and [AgNTf2] (97%, Aldrich, 30.0 mg,
0.075 mmol) were placed in a 2 mL vial. A rubber septum was fitted and
MeCN (1.0 mL) was added. The mixture was magnetically stirred at
room temperature for 30 min observing the precipitation of AgCl. The
liquid was passed through a microfilter syringe and the filtrates were di-
luted to 1,4-dioxane (5 mL) and then 500 mL of this solution was diluted
in MeCN (5 mL) and purged with N2 over 15 min. Cyclic voltammetry
measurements were carried out by using the equipment described in gen-
eral procedures.
[31] A. S. K. Hashmi, Catal. Today 2007, 122, 211.
[32] A. S. K. Hashmi, L. Schwarz, P. Rubenbauer, M. C. Blanco, Adv.
Synth. Catal. 2006, 348, 705.
Acknowledgements
[33] K. S. Williamson, T. P. Yoon, J. Am. Chem. Soc. 2012, 134, 12370.
[34] A. S. K. Hashmi, I. Braun, P. Noesel, J. Schaedlich, M. Wieteck, M.
Rudolph, F. Rominger, Angew. Chem. 2012, 124, 4532; Angew.
Chem. Int. Ed. 2012, 51, 4456.
The work has been supported by Consolider-Ingenio 2010 (proyecto
MULTICAT). J.R.C.A. thanks MCIINN for the concession of a pre-doc-
toral FPU fellowship. A.L.P. thanks ITQ for financial support.
[35] S. Antoniotti, V. Dalla, E. DuÇach, Angew. Chem. 2010, 122, 8032;
Angew. Chem. Int. Ed. 2010, 49, 7860.
[1] W. E. Brenzovich, Jr., Angew. Chem. 2012, 124, 9063; Angew. Chem.
Int. Ed. 2012, 51, 8933.
[2] J. Oliver-Meseguer, J. R. Cabrero-Antonino, A. Leyva-Pꢀrez, I.
Dominguez, A. Corma, Science 2012, 338, 1452.
[3] A. Corma, A. Leyva-Pꢀrez, M. J. Sabater, Chem. Rev. 2011, 111,
1657.
[4] N. Krause, C. Winter, Chem. Rev. 2011, 111, 1994.
[5] H. Huang, Y. Zhou, H. Liu, Beilstein J. Org. Chem. 2011, 7, 897.
[6] A. S. K. Hashmi, Angew. Chem. 2010, 122, 5360; Angew. Chem. Int.
Ed. 2010, 49, 5232.
[7] S. K. Beaumont, G. Kyriakou, R. M. Lambert, J. Am. Chem. Soc.
2010, 132, 12246.
[8] A. S. K. Hashmi, M. Buehrle, Aldrichimica Acta 2010, 43, 27.
[9] N. Marion, R. S. Ramꢆn, S. P. Nolan, J. Am. Chem. Soc. 2009, 131,
448.
[10] A. Grirrane, A. Corma, H. Garcia, Science 2008, 322, 1661.
[11] A. Corma, H. Garcia, Chem. Soc. Rev. 2008, 37, 2096.
[12] C. Gonzꢇlez-Arellano, A. Abad, A. Corma, H. Garcia, M. Iglesias,
F. Sanchez, Angew. Chem. 2007, 119, 1558; Angew. Chem. Int. Ed.
2007, 46, 1536.
[13] A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180.
[14] G. Wienhçfer, F. A. Westerhaus, R. V. Jagadeesh, K. Junge, H.
Junge, M. Beller, Chem. Commun. 2012, 48, 4827.
[15] J. R. Cabrero-Antonino, A. Leyva-Pꢀrez, A. Corma, Adv. Synth.
Catal. 2012, 354, 678.
[36] L. Ricard, F. Gagosz, Organometallics 2007, 26, 4704.
[37] T. T. Dang, F. Boeck, L. Hintermann, J. Org. Chem. 2011, 76, 9353.
[38] J. G. Taylor, L. A. Adrio, K. K. Hii, Dalton Trans. 2010, 39, 1171.
[39] G. Kovꢇcs, A. Lledꢆs, G. Ujaque, Organometallics 2010, 29, 5919.
[40] Z. Li, J. Zhang, C. Brouwer, C.-G. Yang, N. W. Reich, C. He, Org.
Lett. 2006, 8, 4175.
[41] D. C. Rosenfeld, S. Shekhar, A. Takemiya, M. Utsunomiya, J. F.
Hartwig, Org. Lett. 2006, 8, 4179.
[42] T. C. Wabnitz, J.-Q. Yu, J. B. Spencer, Chem. Eur. J. 2004, 10, 484.
[43] J. Penzien, R. Q. Su, T. E. Muller, J. Mol. Catal. A 2002, 182–183,
489.
[44] M. Weiwer, L. Coulombel, E. Dunach, Chem. Commun. 2006, 332.
[45] A. Leyva, A. Corma, J. Org. Chem. 2009, 74, 2067.
[46] A. Leyva, A. Corma, Adv. Synth. Catal. 2009, 351, 2876.
[47] R. Arvai, F. Toulgoat, B. R. Langlois, J.-Y. Sanchez, M. Mꢀdebielle,
Tetrahedron 2009, 65, 5361.
[48] A. S. K. Hashmi, M. C. Blanco, D. Fischer, J. W. Bats, Eur. J. Org.
Chem. 2006, 1387.
[49] N. Morita, N. Krause, Angew. Chem. 2006, 118, 1930; Angew. Chem.
Int. Ed. 2006, 45, 1897.
[50] L. L. Santos, V. R. Ruiz, M. J. Sabater, A. Corma, Tetrahedron 2008,
64, 7902.
[51] T. Hashimoto, S. Kutubi, T. Izumi, A. Rahman, T. Kitamura, J. Or-
ganomet. Chem. 2011, 696, 99.
[52] A. Corma, V. R. Ruiz, A. Leyva-Pꢀrez, M. J. Sabater, Adv. Synth.
Catal. 2010, 352, 1701.
[53] A. S. K. Hashmi, M. Rudolph, Chem. Soc. Rev. 2008, 37, 1766.
[54] A. Leyva-Pꢀrez, A. Corma, Angew. Chem. 2012, 124, 636; Angew.
Chem. Int. Ed. 2012, 51, 614.
[16] J. R. Cabrero-Antonino, A. Leyva-Pꢀrez, A. Corma, Chem. Eur. J.
2012, 18, 11107.
[17] A. Boddien, D. Mellmann, F. Gaertner, R. Jackstell, H. Junge, P. J.
Dyson, G. Laurenczy, R. Ludwig, M. Beller, Science 2011, 333, 1733.
[18] C.-L. Sun, B.-J. Li, Z.-J. Shi, Chem. Rev. 2011, 111, 1293.
[19] K. Junge, K. Schroeder, M. Beller, Chem. Commun. 2011, 47, 4849.
[20] S. Zhou, S. Fleischer, K. Junge, S. Das, D. Addis, M. Beller, Angew.
Chem. 2010, 122, 8298; Angew. Chem. Int. Ed. 2010, 49, 8121.
[21] J. R. Cabrero-Antonino, A. Leyva-Pꢀrez, A. Corma, Adv. Synth.
Catal. 2010, 352, 1571.
[55] D. B. Williams, M. E. Stoll, B. L. Scott, D. A. Costa, W. J. Old-
ham, Jr., Chem. Commun. 2005, 1438.
Received: January 31, 2013
Published online: May 13, 2013
Chem. Eur. J. 2013, 19, 8627 – 8633
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8633