Ruthenium-catalyzed reduction of carbon dioxide to formaldehyde

Article abstract of DOI:10.1021/ja500708w

Functionalization of CO₂ is a challenging goal and precedents exist for the generation of HCOOH, CO, CH₃OH, and CH₄ in mild conditions. In this series, CH₂O, a very reactive molecule, remains an elementary C₁ building block to be observed. Herein we report the direct observation of free formaldehyde from the borane reduction of CO₂ catalyzed by a polyhydride ruthenium complex. Guided by mechanistic studies, we disclose the selective trapping of formaldehyde by in situ condensation with a primary amine into the corresponding imine in very mild conditions. Subsequent hydrolysis into amine and a formalin solution demonstrates for the first time that CO₂ can be used as a C ₁ feedstock to produce formaldehyde.

Full text of DOI:10.1021/ja500708w

Article
pubs.acs.org/JACS
Ruthenium-Catalyzed Reduction of Carbon Dioxide to Formaldehyde
Sebastien Bontemps,* Laure Vendier, and Sylviane Sabo-Etienne*
́
CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne, BP 44099, F-31077 Toulouse Cedex 4, France
́
Universite de Toulouse, UPS, INPT, F-31077 Toulouse Cedex 4, France
S
* Supporting Information
ABSTRACT: Functionalization of CO2 is a challenging goal and
precedents exist for the generation of HCOOH, CO, CH3OH, and
CH4 in mild conditions. In this series, CH2O, a very reactive
molecule, remains an elementary C1 building block to be observed.
Herein we report the direct observation of free formaldehyde from
the borane reduction of CO2 catalyzed by a polyhydride ruthenium
complex. Guided by mechanistic studies, we disclose the selective
trapping of formaldehyde by in situ condensation with a primary
amine into the corresponding imine in very mild conditions.
Subsequent hydrolysis into amine and a formalin solution demonstrates for the first time that CO2 can be used as a C1
feedstock to produce formaldehyde.
noteworthy that at the industrial level, CH2O is a major reactive
C1 source, since more than 20 million tons per year are
produced mainly from methanol oxidation.7 In this context the
synthesis of this versatile molecule would not only be of
fundamental interest but could also expand the scope of
products accessible from the reduction of CO2. For instance,
reduction into formaldehyde and subsequent reaction with a
N−H bond would produce an imine, an important functional
group in chemistry. Related multicomponent strategies were
recently developed for the formylation8 and methylation9 of
N−H bonds.
Recently, we launched a program aimed at studying the
ability of polyhydride ruthenium complexes to catalyze the
reduction of CO2. We first focused on a system based on the
bis(tricyclohexylphosphine) complex [RuH2(H2)2(PCy3)2]
(1Cy) for which borane coordination studies had been
conducted previously in the group.10
INTRODUCTION
■
CO2 is an attractive alternative to fossil resources for the
synthesis of common C1 sources.1 Beyond the CO2 reduction
into formic acid, it has been demonstrated that organometallic
and organic catalytic systems can afford CO, CH3OH, and CH4
under 1 atm of CO2 at room temperature using, in most cases,
boranes or silanes as reductant and oxygen scavenger (Chart
1).2 With dihydrogen as the sole reductant, CH3OH can also be
Chart 1. Mild Catalytic Reduction of CO2 (1 atm) into C1
Building Blocks (corresponding author, resulting oxygen
acceptor product, and year of publication. Reactions mainly
performed at r.t.)
Using pinacolborane as reducing agent and oxygen
scavenger, we reported the fast reduction of CO2 into five
boron compounds, including an original C2 compound
resulting from the unprecedented reductive coupling of two
molecules of CO2.11 An important step in proving the
involvement of formaldehyde was reached when we demon-
strated that this C2 compound, pinBOCH2OCHO (11), was
formed from the reaction of pinBOCHO (9) with transient
CH2O.12
By using the analogous bis(tricyclopentylphosphine) com-
plex [RuH2(H2)2(PCyp3)2] (1Cyp),13 we have shown that a
small modification in the phosphine substituents had a major
impact in catalytic nitrile reduction,14 H/D exchange,15 and
dehydrogenation processes.16 We now disclose the use of 1Cyp
as catalyst precursor for the reduction of carbon dioxide leading
produced through a cascade reaction involving three different
catalysts,3 or by using a tridentate phosphine ruthenium catalyst
precursor under harsher conditions.4 To our knowledge CH2O
is a missing elementary building block that has never been
observed in the catalytic reduction of CO2.5 To account for the
formation of CH3OH, transient formaldehyde has been
postulated and supported by theoretical calculations.6 It is
Received: January 22, 2014
Published: March 10, 2014
© 2014 American Chemical Society
4419
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425

Journal of the American Chemical Society
Article
to the direct observation of free formaldehyde. Mechanistic
considerations on the nature of the active species and on the
role of the C1 and C2 compounds are discussed. Through in situ
condensation of formaldehyde with a primary amine, and
subsequent hydrolysis of the resulting imine into a formalin
solution, we demonstrate the concept of using CO2 as a C1
feedstock to produce formaldehydethe process being
selective under very mild conditions.
with CH2O.12 Here, as a large ratio of formaldehyde is formed,
complete conversion of 9 is achieved. This proves to be the first
direct observation of free formaldehyde from a homogeneously
catalyzed reduction of CO2. Monitoring the mixture over a
longer period shows that final conversion into 7 and 8 is
obtained in a much shorter time, 48 h, by comparison to the 22
days necessary when using 1Cy as catalyst precursor. Addition-
ally, the new C2 compound pinBOCH2OCH3 (12) was
detected as a minor product after 24h (Figure 1).
In an attempt to isolate formaldehyde, we conducted the
standard reaction, and after 30 min, the volatile products were
transferred upon vacuum. NMR spectra of the volatiles showed
formaldehyde as the major compound along with pinBOCH3
(8) and new signals resulting from multiple insertion of
formaldehyde into 8. Among them, pinBOCH2OCH3 (12) and
the new C3 compound pinBOCH2OCH2OCH3 (13) were
characterized. These compounds could also be independently
generated from the reaction of pinBOCH3 and formaldehyde in
the absence of any ruthenium catalyst (Scheme 2).
RESULTS AND DISCUSSION
■
Catalytic Experiments Using 1Cyp as Catalyst Pre-
cursor. Generation of Free Formaldehyde. The dihydride
bis(dihydrogen) complex [RuH2(H2)2(PCyp3)2] (1Cyp) bearing
two tricyclopentylphosphines acts as a catalyst precursor for the
reduction of CO2 mediated by pinacolborane in the same
standard conditions previously reported for complex 1Cy (10
mol %, 1 atm of 13CO2, 30 min, r.t. in a closed NMR tube).11
HBpin is readily consumed, and the different organic
compounds 6−11 are shown in Scheme 1. The relative ratios
To explore further the reduction of CO2, various
experimental conditions have been applied, and the relative
ratios of compounds 6−11 are reported in Table 1. It appears
that selectivity is significantly influenced by the nature of the
catalyst precursor, as well as by the CO2 pressure and the
nature of the solvent. In C6D6 and with a low CO2 pressure
(Table 1, entries 2 and 3), formaldehyde is favored over 9. On
the contrary, in THF or using a higher pressure of CO2 (Table
1, entries 4−6) compound 9 is favored over 6.
Scheme 1. Direct Observation of Free Formaldehyde Using
1Cyp as a Catalyst (see Table 1 for relative ratios)
Mechanistic Studies and Characterization of Com-
plexes 2−4Cyp. To better understand the differences between
the PCy3 and PCyp3 systems leading to the observation of
formaldehyde, we evaluated the fate of the catalyst precursor. In
the cyclohexyl system, we had identified three organometallic
species that played a major role in the catalysis: two
monocarbonyl complexes coexisting all along the catalysis,
RuH(CO2H)(CO)(PCy3)2 (2Cy), and RuH(H2Bpin)(CO)-
(PCy3)2 (4Cy), prior to deactivation into the dicarbonyl
RuH2(CO)2(PCy3)2 (5Cy).11 In the cyclopentyl system, the
Table 1. Relative Ratios of Compounds 6−11, under Various
Conditions (catalyst precursor, CO2 pressure and solvent)
corresponding monocarbonyl complex 2Cyp is readily formed
17
with traces of the inactive dicarbonyl 5Cyp
.
After 24 h, a new
carbonate complex RuH(O2COMe)(CO)(PCyp3)2 (3Cyp) is
observed, whereas the dihydroborate complex 4Cyp was never
relative ratios (%)
6/7/8/9/10/11
catalyst P CO2, (atm) solvent
detected during the catalysis. The new PCyp3 complexes 2Cyp
−
1
2
3
4
5
6
7
8
9
1Cy
1
C6D6
C6D6
C6D6
C6D6
C6D6
THF-d8
C6D6
C6D6
C6D6
0/49/12/13/5/21
22/49/11/0/6/12
18/49/15/0/7/11
0/50/9/6/9/26
0/52/5/13/4/26
0/47/6/10/4/33
1/63/35/0/1/0
1/52/44/0/3/0
1/55/41/0/3/0
4Cyp have been independently synthesized and characterized
(Chart 2). 2Cyp and 4Cyp present NMR data similar to the
corresponding PCy3 complexes. Complex 3Cyp displays a set of
NMR signals presenting strong similarities with complex 2Cyp
1Cyp
1Cyp
1Cyp
1Cyp
1Cyp
2Cyp
3Cyp
4Cyp
1
0.5
1.5
4
2
for the phosphine (δ31P = 45.6 Hz, JP−C = 14.0 Hz), carbonyl
1
(δ13C = 208, t, 2JC−P = 14.0 Hz) and hydride (δ1H = −17.71, td,
1
2
1
2JH−P = 19.8 Hz, JH−C = 11.3 Hz) ligands, but lacking any
signature for a formate ligand. Instead, a set of signals (δ1H =
1
1
3
3.52, dd, JHC = 145.2 Hz, JH−C = 4.0 Hz; δ13C = 158.7 and
52.7) is indicative of a carbonate ligand featuring two 13C-
labeled carbon atoms when 13CO2 is used. X-ray diffraction
analyses have been conducted on complexes 2Cyp, 3Cyp, and
obtained in different conditions are reported in Table 1. One
striking difference between the two systems is in the present
case the absence of the formoxyborane compound 9 (Table 1,
entry 2 vs 1) and the formation in rather large proportions of a
new compound characterized by a resonance at 8.74 ppm in the
1H NMR spectrum, and assigned to formaldehyde 6. This
signal appears as a doublet when 13CO2 is used (1JH−C = 176.6
Hz)providing evidence for its formation from carbon
dioxideand is associated with a carbon NMR signal at
193.0 ppm, indicative of free formaldehyde. We had shown that
the C2 compound 11 could be obtained by direct reaction of 9
4
Cyp (see Supporting Information [SI]). The three compounds
present octahedral arrangements with the two phosphines in
axial position, a hydride (in the case of 3Cyp, statistical disorder
issues prevented hydride location) and a carbonyl in cis
position in the equatorial plane. The carbonyl ligands are
further characterized by IR spectroscopy at 1899, 1894, and
1935 cm−1, respectively. To complete the equatorial plane, the
formate (2Cyp), carbonate (3Cyp), and dihydroborate (4Cyp
)
4420
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425

Journal of the American Chemical Society
Article
1
Figure 1. H NMR stack spectra of the standard reaction with catalyst precursor 1Cyp at various times; 10.5−3.0 ppm region.
5
Cyp. Complex 3Cyp only appeared in small amount after 24h
Scheme 2. Reaction of Formaldehyde with Compounds 8
and 9
(Figure 1) and complex 4Cyp was never detected. We have
independently shown that 3Cyp can be slowly produced from
2Cyp and formaldehyde (Sup. Info. Figure S3). It thus appears
that in the catalytic mixture i.e. in the presence of HBpin and
CO2, complex 2Cyp does not afford complex 4Cyp. However,
when starting from complexes 2Cyp or 3Cyp, introduction of
HBpin readily generates complex 4Cyp. It is thus not surprising
to observe similar catalytic activities when starting from
complexes 2−4Cyp, since 4Cyp is the only complex present in
the system when CO2 is introduced. When the catalysis begins,
complex 4Cyp is readily converted into 2Cyp and 3Cyp in less than
10 min. When starting from 4Cyp, the observation of complex
3Cyp at the early stage of catalysis is in marked contrast with the
standard reaction conducted with 1Cyp (3Cyp observed after 24
h). There is thus a link between the presence of 3Cyp and the
enhanced catalytic activity. Moreover we observed that CO2
addition to isolated 4Cyp resulted in the fast formation of 2Cyp
and 3Cyp in a 1:0.15 ratio and reduction of CO2 into 7 and 8
(SI).
On the basis of the data gained by conducting the CO2
reduction with the PCy3 and PCyp3 systems, we propose the
following mechanism featuring three main catalytic cycles to
account for the accumulation of the different compounds
(Scheme 3). The first step consists in the classical CO2
insertion into a Ru−H bond in competition with HBpin
coordination. This coordination appears to have no impact on
the outcome of the catalysis in the PCy3 system. Guan et al.
also mentioned an innocent behavior of an [Ni]-H·HBCat
adduct in a related CO2 reduction process.2f However, in the
PCyp3 system, the dihydroborate complex 4Cyp may generate
Chart 2. Ruthenium Bis(tricyclopentyl) Complexes 2−5Cyp
ligands are coordinated in a bidentate fashion further
substantiated by stretching frequencies in the case of 2Cyp
(vO2C = 1555 cm−1) and 3Cyp (vO2CO = 1578 cm−1).11,18
Complex 1Cyp exhibits a higher activity than 1Cy; at room
temperature, complete transformation of HBpin and CO2 into
7 and 8 is achieved within 48 h instead of 22 days for 1Cy.
Remarkably, when the standard reaction was conducted with
the isolated complexes 2−4Cyp, compounds 7 and 8 were
obtained in 30 min with only traces amount of 6 and 10 (Table
1, entries 7−9). To explain this enhanced activity, the
differences in the fate of the catalyst precursor were examined
when starting from 1Cyp or from 2−4Cyp. As already mentioned,
when using 1Cyp, complex 2Cyp is readily formed with traces of
4421
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425

Journal of the American Chemical Society
Article
Scheme 3. Proposed mechanism based on the detection of compounds 2−11
the more active carbonate complex 3Cyp. The formate complex
2 then reacts with a first equivalent of HBpin to afford
compound 9, which can further react with the regenerated
catalyst to give rise to an acetal species {[Ru]OCH2OBpin}.
The observation of the acetal compound 10 substantiates the
occurrence of such an intermediate, previously postulated in
borane-based reductive processes.2f,o,6b The reaction of the
acetal complex with a second equivalent of HBpin results in
formal oxygen abstraction and formation of pinBOBpin (7)
along with the release of formaldehyde. Formaldehyde is able to
enter the last cycle to afford a methoxy complex that reacts with
the third equivalent of HBpin and finally releases the
methoxyborane 8. To confirm that the last step is metal-
catalyzed, we conducted two additional experiments; treatment
of HBpin with formaldehyde produced compound 8 in less
than 10 min in the presence of 1Cyp, whereas 8 days were
necessary when no metal catalyst was introduced. In addition,
we have shown that formaldehyde can react in the catalytic
conditions with 2Cyp, 8, and 9 to afford 3Cyp, 12, and 11,
respectively. A similar mechanism involving three cycles for the
reduction of CO2 with HBCat has been proposed by Guan et
al., and further substantiated by Wang et al. calculations.2f,i,o,6b
Our in-depth studies provide conclusive proofs for various steps
of this complex mechanism. Key findings are (i) the direct
observation of formaldehyde and the variation of the
compound relative ratios depending on the conditions, (ii)
the observation of compounds 9, 10, and 11 and their direct
link with formaldehyde,12 (iii) the validation that the reaction
of formaldehyde with HBpin is metal-catalyzed. Moreover, as
observed in previous studies,13,14,16 it is interesting to note the
differences resulting from the use of 1Cyp or 1Cy; conforma-
tional changes in the cycloalkyl rings and solubility properties
being important factors.19
change the relative rates of the three proposed catalytic cycles,
allowing the accumulation and thus the detection of free
formaldehyde. In situ trapping of formaldehyde sounded like an
attractive way to obtain this target molecule and solve
selectivity issues by preventing side reactions occurring at
early stages. The formation of formaldehyde 6 and compound
11 are proofs that trapping is possible. We had shown earlier
that formaldehyde could be recovered from the reaction of
CH3OH with compounds 10 and 11.12 However, when
methanol was added at the beginning of the reaction, it reacted
with HBpin. The more hindered 2,6-bis(diisopropyl)phenol
gave also side reactions before the introduction of CO2. We
then turned our attention to amine functions. Condensation of
ketones or aldehydes with amines is indeed the most common
way to generate imines.20 In addition, these reactions are
reversible, an important requirement to recover and use
formaldehyde. 2,6-Bis(diisopropyl)aniline was chosen for its
protected amine function; it affords the only known stable
monomeric methylene aniline compound upon reaction with
formaldehyde.21 When 2,6-bis(diisopropyl)aniline was intro-
duced in the catalytic system, NMR control prior to CO2
introduction showed that no reaction had occurred. By
applying 1 atm of CO2, we were then pleased to observe the
complete disappearance of HBpin within 1 h and the
appearance of NMR signals associated with methylene aniline
(14) (Scheme 4).21 The methylene moiety presents a
1
characteristic deshielded AB resonance in H NMR at 7.25
1
2
ppm (dd, JH−C = 179.4 Hz, JH−H = 18.5 Hz, 1H, CH2) and
1
2
6.88 ppm (dd, JH−C = 160.5 Hz, JH−H = 18.5 Hz, 1H, CH2)
with a large 1JH−C coupling constant when using labeled 13CO2,
correlating to a carbon signal at 155.2 ppm (Figure S6 of SI).
HRMS analysis on the crude material confirmed the presence
of 12C or 13C labeled-methyleneaniline when 12CO2 and 13CO2
were employed, respectively. To our delight, this reaction is
very selective: HBpin is totally consumed and the 13C NMR
spectrum shows only the signals associated with labeled
compound 14 along with complex 2Cyp (Figure S7 of SI).
Subsequent hydrolysis of the methyleneaniline 14, regenerated
From CO2 Reduction to Formalin: Generation of the
Imine 14. Many efforts have been devoted to better control
selectivity issues in this ruthenium CO2 reduction process. We
have observed that the relative ratios of compounds 6−11 are
sensitive to various conditions. The PCyp3 system appears to
4422
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425

Journal of the American Chemical Society
Article
Varying the solvent (Table 2, entry 3), temperature (Table 2,
entries 4−6), and CO2 pressure (Table 2, entry 7) conditions
have little impact. Moreover, when using complex 1Cy or 4Cyp as
catalyst precursors, compound 14 was also obtained in
moderate to good yield based on HBpin (Table 2, entries 8−
11). Finally, one major drawback of the reduction process in
the absence of aniline was the catalyst deactivation that
prevented us from decreasing the catalyst loading. Trapping
formaldehyde appears to suppress the deactivation process
since the catalyst loading can be decreased down to 0.5 mol %
with 54% yield after 5 h (Table 2, entries 12−14), the active
catalyst 2Cyp being the only complex detected at the end of the
reaction. As a summary, by preventing competitive reactions in
this intricate system, trapping formaldehyde in situ successfully
allowed the establishment of a catalyzed CO2 functionalization
process exhibiting high selectivity with low catalyst loading.
Scheme 4. From CO2 Reduction to Formalin: Generation of
the Imine 14 and Subsequent Hydrolysis to Formalin and
Amine
aniline and afforded a formalin solution (Figures S8 and S9 of
SI), thus demonstrating the concept of using CO2 as a C1
feedstock to produce formaldehyde.
The yield in methylene aniline 14, based on 2 equiv of
HBpin for the transformation of CO2 into CH2O, was
CONCLUSION
1
■
determined by H NMR and reported in Table 2. One can
Formaldehyde was an elementary C1 building block that had
never been observed in the homogeneous reduction of CO2. By
using the bis(dihydrogen) complex [RuH2(H2)2(PCyp3)2]
(1Cyp) we describe the catalyzed reduction of CO2 into various
organic compounds including free formaldehyde. Guided by in-
depth mechanistic studies, we were able to selectively trap
formaldehyde by in situ condensation to a primary amine
affording the corresponding imine under very mild conditions.
Subsequent hydrolysis provided a formalin solution and
regenerated the amine, thus demonstrating for the first time
that CO2 can be used as a C1 feedstock to produce
formaldehyde. Further research to optimize this concept is
currently underway in our group.
Table 2. Variation of Catalytic Conditions and
Corresponding Yields in Compound 14
catalyst
loading %
P (CO2)
(atm)
T
yield in 14
a
entry
solvents (°C)
(%)
1
1Cyp (10)
1Cyp (7)
1Cyp (10)
1Cyp (1)
1Cyp (10)
1Cyp (1)
1Cyp (10)
1Cy (10)
4Cyp (2)
4Cyp (1)
4Cyp (7)
1Cyp (5)
1Cyp (1)
1Cyp (0.5)
1
1
1
1
1
1
3
1
1
1
1
1
1
1
C6D6
C6D6
THF-D8
C6D6
C6D6
C6D6
C6D6
C6D6
C6D6
C6D6
C6D6
C6D6
C6D6
C6D6
25
25
25
10
70
70
25
25
25
25
25
25
25
25
47
74
42
49
32
43
36
29
51
50
74
47
50
54
b
2
3
c
4
5
c
6
7
8
9
c
EXPERIMENTAL SECTION
10
■
b
11
General Methods. All reactions and manipulations were
carried out under an atmosphere of dry argon using standard
Schlenk or glovebox techniques. Solvents were dried using an
MBraun SPS column. Deuterated benzene, toluene, and THF
were freeze−pump−thaw degassed and stored under Ar over 4
Å molecular sieves. Quick pressure valve NMR tubes were used
12
c
13
d
14
a
Catalyst loading and yield based on HBpin, unless otherwise stated.
Catalyst loading and yield based on CO2. 3 h of reaction. 5 h of
b
c
d
1
for reactions with CO2. H, 13C, 11B, and 31P NMR spectra
reaction.
were recorded on Bruker AV 400 or 500 spectrometers.
Chemical shifts are expressed with a positive sign, in parts per
be surprised by the observed 47% yield as 14 was the only
compound detected (Table 2, entry 1). A rather simple
explanation is that HBpin is involved not only as a reductant of
carbon dioxide but also as a dehydrating agent, driving the
reaction toward the formation of the imine by trapping water
formed during the condensation of aniline with formaldehyde.
Indeed, 1−2 equiv of HBpin reacted with water to generate
pinBOBpin, pinBOH, and H2 as observed by NMR. Thus, the
maximum yield based on our calculation ranges formally
between 50 and 66%.
To account for the selectivity observed by NMR spectros-
copy and calculate a yield based on CO2, we conducted the
catalysis by transferring a known amount of CO2 to an excess of
HBpin and aniline with 7 mol % catalyst (Table 2, entry 2). If 2
equiv of CO2 reacted with 1Cyp to generate 2Cyp, thus limiting
the yield at a 86% maximum, the measured 74% yield after 1 h
corroborates that 14 and complex 2Cyp were solely detected in
the 13C{1H} NMR spectrum with no trace of free 13CO2. We
then applied various catalytic conditions and observed minor
differences, as opposed to the catalytic system without aniline.
1
million, relative to residual H and 13C solvent signals, and
external BF3·OEt2 and 85% H3PO4. Complexes 1Cyp and 5Cyp
were synthesized according to literature procedures.13,17
Synthesis and Characterization of 2−4Cyp. 2Cyp: A
solution of Ru(H)2(η2-H2)(CO)(PCyp3)2 (50 mg, 0.08 mmol)
in Et2O (2 mL) was exposed to 1 atm of CO2 for 15 min at
ambient temperature. The solvent was removed by filtration,
and the residue was washed twice with pentane (0.5 mL) at
−30 °C, affording the expected complex Ru(H)(O2CH)(CO)-
(PCyp3)2 as an orange solid in 53% yield. Crystals suitable for
X-ray analysis were grown from the slow evaporation of a
1
pentane solution at room temperature. Mp = 163−165 °C. H
NMR (400.1 MHz, C6D6, 298 K): δ 8.20 (s, 1H, O2CH),
2
2.25−1.45 (m, 54H, Cyp), −17.83 (t, JH−P = 19.6 Hz, 1H,
Ru−H); 13C{1H} NMR (100.6 MHz, C6D6, 298 K): δ 207.9 (t,
2JC−P = 13.9 Hz, CO), 172.9 (s, O2CH), 37.4 (t, JC−P = 11.5 Hz,
Cyp), 30.3 (s, Cyp), 29.9 (s, Cyp), 26.5 (t, JC−P = 4.0 Hz, Cyp),
26.3 (t, JC−P = 4.0 Hz, Cyp); 31P{1H} NMR (162.0 MHz, C6D6,
298 K): δ 45.8. IR (solid, cm−1): 2040 (weak, vRuH), 1899
4423
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425

Journal of the American Chemical Society
Article
3
(very strong, vCO), 1555 (strong, vO2C). Anal. Cald. for
C32H57O3P2Ru: C, 58.88; H, 8.80. Found: C, 59.12; H, 8.48.
Ru(H)(O213CH)(13CO)(PCyp3)2 has been characterized in
situ in the standard reaction. 1H NMR (400.1 MHz, C6D6, 298
Hz, JH−C = 4.8 Hz, 3H, CH3); 13C{1H} NMR (100.6 MHz,
C6D6, 298 K): δ 94.0 (pseudo-t, 2JC−C = 2.3 Hz, CH2), 87.2 (d,
2JC−C = 2.2 Hz, CH2), 55.2 (d, JC−C = 2.3 Hz, CH3).
2
14: 1H NMR (400.1 MHz, C6D6, 298 K): δ 7.25 (dd, 1JH−C
=
1
2
K): δ 8.20 (d, JH−C = 195.5 Hz, 1H, O2CH), 2.25−1.45 (m,
179.4 Hz, JH−H = 18.5 Hz, 1H, CH2), 7.13 − 7.02 (m, 3H,
CH(arom)), 6.88 (dd, 1JH−C = 160.5 Hz, 2JH−H = 18.5 Hz, 1H,
CH2), 2.97 (hept, 3JH−H = 6.9 Hz, 2H, CH(i-Pr), 1.12 (d, 3JH−H
= 6.9 Hz, 12H, CH3(i-Pr)); 13C{1H} NMR (100.6 MHz, C6D6,
54H, Cyp), −17.84 (td, 2JH−P = 19.6 Hz, 2JH−C = 11.4 Hz, 1H,
Ru−H); 31P{1H} NMR (162.0 MHz, C6D6, 298 K): δ 45.8 (d,
2JP−C = 13.9 Hz)
298 K): δ 155.2 (s, 13CH2).
3Cyp: A solution of Ru(H)(O2CH)(CO)(PCyp3)2 (150 mg,
0.20 mmol) in toluene/methanol (2 mL/2 mL) was stirred at
70 °C for 6 h. The solvents were removed by vacuum, and the
residue was washed with methanol (3*2 mL), affording
Ru(H)(O2COCH3)(CO)(PCyp3)2 in 74% yield. Monocrystals
suitable for X-ray analysis were obtained from a concentrated
solution of diethylether or toluene/methanol at room temper-
ature. In each case, poor definition resulting from statistical
disorder did not allow any discussion about angles and bond
1H NMR (400.1 MHz, Tol-d8, 298 K): δ 7.30 (dd, JH−C
=
1
2
179.5 Hz, JH−H = 18.6 Hz, 1H, CH2), 7.16−7.02 (m, 3H,
CH(arom)), 6.92 (dd, 1JH−C = 160.5 Hz, 2JH−H = 18.5 Hz, 1H,
CH2), 3.38 (hept, 3JH−H = 6.9 Hz, 2H, CH(i-Pr), 1.23 (d, 3JH−H
= 6.9 Hz, 12H, CH3(i-Pr)); 13C{1H} NMR (100.6 MHz, C6D6,
298 K): δ 155.4 (s, 13CH2); HRMS (DCI-CH4): m/z (M+:
C H19N): calculated: 189.1517, found: 189.1522; m/z (M+:
12
13
12
C
13CH19N): calculated: 190.1566, found: 190.1551.
12
1
1H NMR (400.1 MHz, THF-d8, 298 K): δ 7.73 (dd, 1JH−C
=
distances (see SI). Mp = 142−144 °C. H NMR (400.1 MHz,
179.5 Hz, JH−H = 18.4 Hz, 1H, CH2), 7.37 (dd, 1JH−C = 160.5
2
C6D6, 298 K): δ 3.53 (s, 3H, CH3), 2.25−1.44 (m, 54H, Cyp),
−17.68 (t, 2JH−P = 19.6 Hz, 1H, Ru−H); 13C{1H} NMR (100.6
MHz, C6D6, 298 K): δ 208.0 (t, 2JC−P = 14.0 Hz, CO), 158.7 (s,
O2COCH3), 52.7 (s, CH3), 37.3 (t, JC−P = 11.3 Hz, Cyp), 30.2
(s, Cyp), 29.8 (s, Cyp), 26.5 (t, JC−P = 4.2 Hz, Cyp), 26.4 (t,
JC−P = 4.2 Hz, Cyp); 31P{1H} NMR (162.0 MHz, C6D6, 298
K): δ 45.6. IR (solid, cm−1): 1894 (very strong, vCO), 1578
(strong, vO2C). Anal. Cald. for C33H59O4P2Ru: C,58.05;
H,8.71. Found: C,57.45; H,8.64
2
Hz, JH−H = 18.4 Hz, 1H, CH2), 7.13 − 6.97 (m, 3H,
CH(arom)), 2.91 (hept, 3JH−H = 6.9 Hz, 2H, CH(i-Pr), 1.16 (d,
3JH−H = 6.9 Hz, 12H, CH3(i-Pr)); 13C{1H} NMR (100.6 MHz,
C6D6, 298 K): δ 156.5 (s, 13CH2).
Catalytic Experiments. Generation of Compounds 6−
11. In a pressurizable NMR tube, a solution of a ruthenium
complex (0.013 mmol) and HBpin (17 mg, 0.130 mmol) was
degassed and placed under a pressure of 13CO2 at room
temperature. NMR characterization was conducted after 30 min
Ru(H)(O213CO13CH3)(13CO)(PCyp3)2 has been character-
1
showing the formation of 6−11.11 For 13CH2O: H NMR
1
ized in situ in the standard reaction. H NMR (400.1 MHz,
1
3
(400.1 MHz, C6D6, 298 K) δ = 8.74 (d, 1JH−C = 176.6 Hz, 1H);
13C{1H}, NMR (100.6 MHz, C6D6, 298 K): δ = 193.0 (s).
Generation of Compound 14. In a pressurizable NMR tube,
a solution of a ruthenium complex, HBpin (17 mg, 0.13 mmol),
and 2,6-bis(diisopropyl)aniline (18 mg, 0.10 mmol) was
degassed and placed under a pressure of 13CO2. Catalyst
loading and yields given in Table 2 were based on HBpin,
considering that 2 equiv of HBpin were necessary to reduce
CO2 to CH2O, except otherwise stated. For entries 2 and 11, a
known quantity of 13CO2 (0.19 mmol) was vacuum transferred
to a J-Young tube containing HBpin (99 mg, 0.77 mmol), 2,6-
bis(diisopropyl)aniline (60 mg, 0.34 mmol) and a ruthenium
catalyst (0.013 mmol) in C6D6 solution; the yields were then
calculated based on CO2. The yields were determined by NMR
by using a known quantity of 4-methyl-anisole as a standard,
added to the tube at the end of the reaction.
C6D6, 298 K): δ 3.52 (dd, JH−C = 145.2 Hz, JH−C = 4.0 Hz,
3H, CH3), 2.25−1.44 (m, 54H, Cyp), −17.71 (td, 2JH−P = 19.8
Hz, 2JH−C = 11.3 Hz, 1H, Ru−H); 31P{1H} NMR (162.0 MHz,
2
C6D6, 298 K): δ 45.6 (d, JP−C = 14.0 Hz).
4
Cyp: HBpin (50 mg, 0.39 mmol) was added in excess to a
solution of Ru(H)2(H2)(CO)(PCyp3)2 (89 mg, 0.15 mmol) in
pentane (5 mL) and stirred for 3 h. The solvent and excess
HBpin were then removed by vacuum, and the residue was
washed with cold pentane, affording complex 4Cyp in a 48%
yield. Monocrystals suitable for X-ray analysis were obtained
from a concentrated solution of pentane at room temperature.
1
Mp = 76−78 °C. H NMR (C6D6, 298 K, 500.3 MHz) δ =
2.46−2.38 (m, 6H, Cyp), 2.05−1.98 (m, 12H, Cyp), 1.93−1.83
(m, 12H, Cyp), 1.77−1.69 (m, 12H, Cyp), 1.61−1.48 (m, 12H,
Cyp), 1.16 (s, 12H, CH3(pin)), −6.90 (br, 1H, Ru−H), −8.81
(br, 1H, Ru−H), −9.76 (br, 1H, Ru−H); 13C{1H} NMR
(C6D6, 298 K, 125.8 MHz) δ = 208.7 (t, 2JC−P = 12.0 Hz, CO),
81.5 (s, C(pin)), 39.7 (pseudo-t, JC−P = 12.6 Hz, Cipso (Cyp)),
30.1 (s, Cyp), 30.0 (s, Cyp), 26.4 (s, 2C, Cyp), 24.5 (s,
CH3(pin)); 11B{1H} NMR (C6D6, 298 K, 128.4 MHz) δ =
29.9 (br w1/2 = 265 Hz); 31P{1H} NMR (C6D6, 298 K, 162.0
MHz) δ = 56.2. IR (solid, cm−1): 1935 (strong, vCO). Any
attempt to characterize the compound by elemental analysis or
mass spectrometry failed, presumably because of the poor
stability of the complex observed upon storage in the glovebox
at room temperature.
ASSOCIATED CONTENT
* Supporting Information
Experimental spectroscopic and X-ray crystallographic data for
2−4Cyp. This material is available free of charge via the Internet
at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
■
Corresponding Authors
bontemps@lcc-toulouse.fr
sylviane.sabo@lcc-toulouse.fr
1
In situ characterization of 12−14. 12: H NMR (400.1
1
3
MHz, C6D6, 298 K) δ = 4.99 (dd, JH−C = 164.6, JH−C = 7.0,
2H), 3.23 (dd, 1JH−C = 141.9, 3JH−C = 4.9, 3H); 13C{1H} NMR
(100.6 MHz, C6D6, 298 K): δ = 91.7 (d, JC−C = 2.7, CH2),
Notes
2
The authors declare no competing financial interest.
2
55.6 (d, JC−C = 2.7, CH3).
13: 1H NMR (400.1 MHz, C6D6, 298 K): δ 5.19 (dd, 1JH−C
169.9 Hz, JH−C = 6.1 Hz, 2H, CH2), 4.68 (pseudo-td, JH−C
163.6 Hz, JH−C = 6.0 Hz, 2H, CH2), 3.12 (dd, JH−C = 141.3
=
=
ACKNOWLEDGMENTS
Dr. Vincent Ces
discussion on formaldehyde trapping. S.B. and S.S E. thank
■
3
1
́
ar is warmly acknowledged for helpful
3
1
4424
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425

Journal of the American Chemical Society
Article
(11) Bontemps, S.; Vendier, L.; Sabo-Etienne, S. Angew. Chem., Int.
Ed. 2012, 51, 1671.
(12) Bontemps, S.; Sabo-Etienne, S. Angew. Chem., Int. Ed. 2013, 52,
the ANR (Programme blanc “IRONHYC” ANR-12), the
CNRS for support and Johnson Matthey plc for the generous
gift of hydrated ruthenium trichloride.
10253.
(13) Grellier, M.; Vendier, L.; Chaudret, B.; Albinati, A.; Rizzato, S.;
Mason, S.; Sabo-Etienne, S. J. Am. Chem. Soc. 2005, 127, 17592.
(14) Reguillo, R.; Grellier, M.; Vautravers, N.; Vendier, L.; Sabo-
Etienne, S. J. Am. Chem. Soc. 2010, 132, 7854.
(15) Grellier, M.; Mason, S. A.; Albinati, A.; Capelli, S. C.; Rizzato, S.;
Bijani, C.; Coppel, Y.; Sabo-Etienne, S. Inorg. Chem. 2012, 52, 7329.
(16) Grellier, M.; Vendier, L.; Sabo-Etienne, S. Angew. Chem., Int. Ed.
2007, 46, 2613.
REFERENCES
■
(1) (a) Arakawa, H.; Aresta, M.; Armor, J. N.; Barteau, M. A.;
Beckman, E. J.; Bell, A. T.; Bercaw, J. E.; Creutz, C.; Dinjus, E.; Dixon,
D. A.; Domen, K.; DuBois, D. L.; Eckert, J.; Fujita, E.; Gibson, D. H.;
Goddard, W. A.; Goodman, D. W.; Keller, J.; Kubas, G. J.; Kung, H.
H.; Lyons, J. E.; Manzer, L. E.; Marks, T. J.; Morokuma, K.; Nicholas,
K. M.; Periana, R.; Que, L.; Rostrup-Nielson, J.; Sachtler, W. M. H.;
Schmidt, L. D.; Sen, A.; Somorjai, G. A.; Stair, P. C.; Stults, B. R.;
Tumas, W. Chem. Rev. 2001, 101, 953. (b) Jessop, P. G.; Joo, F.; Tai,
C.-C. Coord. Chem. Rev. 2004, 248, 2425. (c) Sakakura, T.; Choi, J.-C.;
Yasuda, H. Chem. Rev. 2007, 107, 2365. (d) Aresta, M.; Dibenedetto,
A. Dalton Trans. 2007, 2975. (e) Benson, E. E.; Kubiak, C. P.;
Sathrum, A. J.; Smieja, J. M. Chem. Soc. Rev. 2009, 38, 89. (f) Riduan,
S. N.; Zhang, Y. Dalton Trans. 2010, 39, 3347. (g) Wang, W.; Wang,
S.; Ma, X.; Gong, J. Chem. Soc. Rev. 2011, 40, 3703. (h) Cokoja, M.;
(17) Bolton, P. D.; Grellier, M.; Vautravers, N.; Vendier, L.; Sabo-
Etienne, S. Organometallics 2008, 27, 5088.
(18) Burling, S.; Kociok-Kohn, G.; Mahon, M. F.; Whittlesey, M. K.;
̈
Williams, J. M. J. Organometallics 2005, 24, 5868.
(19) Grellier, M.; Sabo-Etienne, S. Chem. Commun. 2012, 48, 34.
(20) Layer, R. W. Chem. Rev. 1963, 63, 489.
(21) (a) Verardo, G.; Cauci, S.; Giumanini, A. G. J. Chem. Soc., Chem.
Commun. 1985, 0, 1787. (b) Giumanini, A. G.; Verardo, G.; Poiana, M.
J. Prakt. Chem./Chem.-Ztg. 1988, 330, 161.
Bruckmeier, C.; Rieger, B.; Herrmann, W. A.; Kuhn, F. E. Angew.
̈
Chem., Int. Ed. 2011, 50, 8510. (i) Fan, T.; Chen, X.; Lin, Z. Chem.
Commun. 2012, 48, 10808. (j) Omae, I. Coord. Chem. Rev. 2012, 256,
1384.
(2) (a) Laitar, D. S.; Mueller, P.; Sadighi, J. P. J. Am. Chem. Soc. 2005,
127, 17196. (b) Matsuo, T.; Kawaguchi, H. J. Am. Chem. Soc. 2006,
128, 12362. (c) Gu, L.; Zhang, Y. J. Am. Chem. Soc. 2009, 132, 914.
(d) Riduan, S. N.; Zhang, Y.; Ying, J. Y. Angew. Chem., Int. Ed. 2009,
48, 3322. (e) Ashley, A. E.; Thompson, A. L.; O’Hare, D. Angew.
Chem., Int. Ed. 2009, 48, 9839. (f) Chakraborty, S.; Zhang, J.; Krause, J.
A.; Guan, H. J. Am. Chem. Soc. 2010, 132, 8872. (g) Berkefeld, A.;
Piers, W. E.; Parvez, M. J. Am. Chem. Soc. 2010, 132, 10660.
(h) Kleeberg, C.; Cheung, M. S.; Lin, Z.; Marder, T. B. J. Am. Chem.
Soc. 2011, 133, 19060. (i) Chakraborty, S.; Patel, Y. J.; Krause, J. A.;
Guan, H. Polyhedron 2012, 32, 30. (j) Sgro, M. J.; Stephan, D. W.
́
Angew. Chem., Int. Ed. 2012, 51, 11343. (k) Park, S.; Bezier, D.;
Brookhart, M. J. Am. Chem. Soc. 2012, 134, 11404. (l) Mitton, S. J.;
Turculet, L. Chem.Eur. J. 2012, 18, 15258. (m) Berkefeld, A.; Piers,
W. E.; Parvez, M.; Castro, L.; Maron, L.; Eisenstein, O. Chem. Sci.
2013, 4, 2152. (n) Dobrovetsky, R.; Stephan, D. W. Angew. Chem., Int.
Ed. 2013, 52, 2516. (o) Chakraborty, S.; Zhang, J.; Patel, Y. J.; Krause,
J. A.; Guan, H. Inorg. Chem. 2013, 52, 37. (p) Courtemanche, M.-A.;
Leg
9326.
́
are,
́
M.-A.; Maron, L.; Fontaine, F.-G. J. Am. Chem. Soc. 2013, 135,
(3) Huff, C. A.; Sanford, M. S. J. Am. Chem. Soc. 2011, 133, 18122.
(4) Wesselbaum, S.; vom Stein, T.; Klankermayer, J.; Leitner, W.
Angew. Chem., Int. Ed. 2012, 51, 7499.
(5) For stoichiometric reaction see: Gambarotta, S.; Strologo, S.;
Floriani, C.; Chiesi-Villa, A.; Guastini, C. J. Am. Chem. Soc. 1985, 107,
6278.
(6) (a) Huang, F.; Lu, G.; Zhao, L.; Li, H.; Wang, Z.-X. J. Am. Chem.
Soc. 2010, 132, 12388. (b) Huang, F.; Zhang, C.; Jiang, J.; Wang, Z.-X.;
Guan, H. Inorg. Chem. 2011, 50, 3816.
(7) (a) Wang, Z.-C.; Dietl, N.; Kretschmer, R.; Ma, J.-B.; Weiske, T.;
Schlangen, M.; Schwarz, H. Angew. Chem., Int. Ed. 2012, 51, 3703.
(b) Salthammer, T. Angew. Chem., Int. Ed. 2013, 52, 3320.
́
(8) (a) Das Neves Gomes, C.; Jacquet, O.; Villiers, C.; Thuery, P.;
Ephritikhine, M.; Cantat, T. Angew. Chem., Int. Ed. 2012, 51, 187.
(b) Jacquet, O.; Das Neves Gomes, C.; Ephritikhine, M.; Cantat, T. J.
Am. Chem. Soc. 2012, 134, 2934. (c) Jacquet, O.; Das Neves Gomes,
C.; Ephritikhine, M.; Cantat, T. ChemCatChem 2013, 5, 117.
(9) (a) Jacquet, O.; Frogneux, X.; Das Neves Gomes, C.; Cantat, T.
Chem. Sci. 2013, 4, 2127. (b) Li, Y.; Fang, X.; Junge, K.; Beller, M.
Angew. Chem., Int. Ed. 2013, 52, 9568. (c) Li, Y.; Sorribes, I.; Yan, T.;
Junge, K.; Beller, M. Angew. Chem., Int. Ed. 2013, 52, 12156.
(10) (a) Montiel-Palma, V.; Lumbierres, M.; Donnadieu, B.; Sabo-
Etienne, S.; Chaudret, B. J. Am. Chem. Soc. 2002, 124, 5624.
(b) Lachaize, S.; Essalah, K.; Montiel-Palma, V.; Vendier, L.; Chaudret,
B.; Barthelat, J.-C.; Sabo-Etienne, S. Organometallics 2005, 24, 2935.
4425
dx.doi.org/10.1021/ja500708w | J. Am. Chem. Soc. 2014, 136, 4419−4425