Combining Low-Pressure CO2 Capture and Hydrogenation to Form Methanol

Article abstract of DOI:10.1021/acscatal.5b00194

This paper describes a novel approach to CO₂ hydrogenation, in which CO₂ capture with aminoethanols at low pressure is coupled with hydrogenation of the captured product, oxazolidinone, directly to MeOH. In particular, (2-methylamino)ethanol or valinol captures CO₂ at 1-3 bar in the presence of catalytic Cs₂CO₃ to give the corresponding oxazolidinones in up to 65-70 and 90-95% yields, respectively. Efficient hydrogenation of oxazolidinones was achieved using PNN pincer Ru catalysts to give the corresponding aminoethanol (up to 95-100% yield) and MeOH (up to 78-92% yield). We also have shown that both CO₂ capture and oxazolidinone hydrogenation can be performed in the same reaction mixture using a simple protocol that avoids intermediate isolation or purification steps. For example, CO₂ can be captured by valinol at 1 bar with Cs₂CO₃ catalyst followed by 4-isopropyl-2-oxazolidinone hydrogenation in the presence of a bipy-based pincer Ru catalyst to produce MeOH in 50% yield after two steps.

Full text of DOI:10.1021/acscatal.5b00194

Research Article
pubs.acs.org/acscatalysis
Combining Low-Pressure CO Capture and Hydrogenation To Form
2
Methanol
†
Julia R. Khusnutdinova, Jai Anand Garg, and David Milstein*
Department of Organic Chemistry, Weizmann Institute of Science, Rehovot 76100, Israel
*
S Supporting Information
ABSTRACT: This paper describes a novel approach to CO2
hydrogenation, in which CO capture with aminoethanols at
2
low pressure is coupled with hydrogenation of the captured
product, oxazolidinone, directly to MeOH. In particular, (2-
methylamino)ethanol or valinol captures CO at 1−3 bar in the
2
presence of catalytic Cs CO to give the corresponding
2
3
oxazolidinones in up to 65−70 and 90−95% yields,
respectively. Efficient hydrogenation of oxazolidinones was
achieved using PNN pincer Ru catalysts to give the
corresponding aminoethanol (up to 95−100% yield) and
MeOH (up to 78−92% yield). We also have shown that both
CO capture and oxazolidinone hydrogenation can be performed in the same reaction mixture using a simple protocol that avoids
2
intermediate isolation or purification steps. For example, CO can be captured by valinol at 1 bar with Cs CO catalyst followed
2
2
3
by 4-isopropyl-2-oxazolidinone hydrogenation in the presence of a bipy-based pincer Ru catalyst to produce MeOH in 50% yield
after two steps.
KEYWORDS: carbon dioxide, hydrogenation, methanol, oxazolidinone, ethanolamine, ruthenium, pincer
INTRODUCTION
Scheme 1. Chemical Processes That Occur during CO
Capture with Amines
2
■
Capturing and converting CO gas to a liquid fuel is an
2
important challenge for the future of a sustainable economy. In
particular, the reaction of CO with H , the latter obtained via
2
2
renewable energy methods such as solar-driven water splitting,
to form a liquid methanol fuel is an important pathway outlined
1
in “The Methanol Economy”. Because the integration of new
energy sources into developed economies must necessarily
proceed in a sequential manner, recent interest has focused on
utilizing CO2 streams from readily available power plant
discharge or from natural gas streams, with a view toward
transforming this waste gas to methanol fuel by reacting it with
One of the major problems of such processes is thermal
degradation of amines occurring during thermal treatment of
the hydrocarbonate and carbamate salts to liberate CO and is
2
2
responsible for the large amount of amine waste resulting from
hydrogen in an energy-efficient manner. Many efforts of power
such CO capture operations. In 2009, for example, Bellona
utilities today are devoted to CO capture and storage, meaning
2
2
Foundation reported that “a typical CO capture plant with the
a recycling option via converting waste CO to methanol fuel is
2
2
capacity of 1 million tonnes of CO annually is expected to
particularly attractive.
Various technologies have been developed for the
2
produce from 300 to 3000 tonnes of amine waste annually”,
9
thereby reducing the economic viability of such processes. The
minimization of CO emissions to the atmosphere from the
2
most common reported thermal degradation product from CO2
capture with monoethanolamine is 2-oxazolidinone (2 in
flue gas streams of fossil-fuel-powered plants. Most common
methods are based on chemical CO absorption using amines
2
Scheme 2), which further reacts to give other products of
or aminoalcohols as solvents, with monoethanolamine (MEA,
10
3
decomposition (3, 4, and 5 in Scheme 2). An additional
1
) being the most common CO capture agent (Scheme 1).
2
energy penalty comes from the need to heat the carbonate salt
Current developed technologies for CO2 capture involve
solutions to high temperatures in order to release CO and
several steps, including CO absorption to produce hydro-
2
2
regenerate the amine and also from the subsequent CO2
carbonate (eq 1, Scheme 1) and carbamate (eq 2, Scheme 1)
4
,5
salts, followed by subsequent release of CO by the high-
2
6,7
temperature decomposition of these salts. The CO gas
Received: January 30, 2015
Revised: March 5, 2015
2
separated by such methods is then compressed and transported
6
,8
to a storage site, while amine solutions are partially recycled.
©
XXXX American Chemical Society
2416
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422

ACS Catalysis
Research Article
Scheme 2. Byproducts of CO Capture with
Chart 1. Ru Hydrogenation Catalysts Used in This Study
2
Monoethanolamine 1 [2 = 2-Oxazolidinone, 3 = N,N′-(2-
Hydroxyethyl)urea, 4 = N-(2-Hydroxyethyl)ethylenediamine
(
(
HEEDA); 5 = 1-(2-Hydroxyethyl)-2-imidazolidone
HEIA)]
12−15
making such processes less attractive.
By contrast, the
approach presented here allows for CO capture at low pressure
2
and direct utilization of the formed captured species for the
generation of a valuable product: MeOH. It appears that, while
this article was under preparation, a similar approach was
recently reported by Sanford et al. based on the use of
dimethylamine for hydrogenation of CO to formamide and
2
16
MeOH. For comparison, the previously reported system for
8
sequestration. Therefore, it would be highly beneficial if,
CO capture in an amidine base/alcohol mixture failed to
2
instead of releasing and storing the CO gas, the products of
2
produce MeOH upon catalytic hydrogenation, and only
CO chemical absorption with amines were to be directly
17
2
formate ester and formate salt were formed.
converted into a liquid fuel such as methanol.
Taking a page from industry, where aminoalcohols are used
RESULTS AND DISCUSSION
■
to fix CO streams and thus “capture” the gas, we hereby report
2
CO Capture. CO capture using monoethanolamine and
2
2
a procedure where we capture CO at low pressure and use the
2
other aminoalcohols has been studied previously and proceeds
through the formation of hydrocarbonate salts as well as
carbamate salts with a protonated amine as the counterion.
During thermal decomposition of these salts that takes place
during the CO2 liberation step in CO2 capture plants,
oxazolidinone 2 (Scheme 2) is formed as a byproduct which
further reacts to give 3, 4, and 5 (Scheme 2). However, the
formation of oxazolidinone can be made selective through the
in situ created capture product, an oxazolidinone compound, to
generate methanol in one step in yields up to a total of ∼50%
Scheme 3). These results represent a novel approach for the
(
Scheme 3. Proposed Scheme for Selective CO Capture and
2
Conversion to MeOH
n
18
19,20
use of catalysts such as Bu SnO, CeO2,
or other systems.
2
Considering that CO capture from gas streams should occur at
2
low partial CO pressures, we focused our attention on a
2
recently reported method for selective oxazolidinone formation
from aminoalcohols and 1 atm of CO using a Cs CO catalyst
2
2
3
21
in dimethylsulfoxide (DMSO) solvent developed by Saito.
According to Saito’s density functional theory results,
monoethanolamine (1) was not an ideal substrate for such
cyclization due to unfavorable thermodynamics of the reaction,
which leads to low yields when only 1 atm of CO is used.
2
use of CO capture products directly for their conversion to
MeOH, thus avoiding additional energy costs required for CO2
thermal regeneration and storage. This approach may be of
However, substrates with bulky substituents in the α-position
to the amine group, such as valinol 6 (derived from the natural
amino acid, valine), show a higher propensity toward
cyclization and form the corresponding oxazolidinone product
2
practical use to the power utility industry which uses a similar
9
21
chemical process to capture CO waste gas.
in up to 90% yield at a pressure of 1 atm of CO2.
2
In the proposed approach, the selective formation of
Indeed, when monoethanolamine 1 was reacted under 2 bar
oxazolidinone via CO capture at low pressures is accomplished
of CO in the presence of 10 mol % of Cs CO in DMSO for
2
2
2
3
through the use of Cs CO as a catalyst, while hydrogenation of
90 h at 150 °C, only 19% of 2-oxazolidinone 2 was obtained
(Scheme 4a). By contrast, valinol 6 reacted much more readily
to produce the corresponding 4-isopropyl-2-oxazolidinone 7 in
90−95% yield after heating at 150 °C for 24 h under 1 bar of
2
3
the latter to produce MeOH and regenerated aminoalcohol is
achieved in the same reaction mixture using Ru pincer catalysts
developed in our group (Scheme 3 and Chart 1).
This approach is conceptually different from the currently
CO (Scheme 4b). A control experiment using 1 mmol of
2
known examples for direct hydrogenation of CO to MeOH
using transition metal catalysts. In particular, the existing
Cs CO and 1 mmol of valinol in the absence of CO gas does
2
2
3
2
11
not generate the oxazolidinone product, showing that Cs CO
2
3
homogeneous catalytic systems for direct CO hydrogenation
does not act as a source of CO in this reaction.
We have also examined the reactivity of 2-(methylamino)-
ethanol 8 under these conditions. 8 was proposed previously as
2
2
to MeOH rely on the use of pressurized CO gas (10−20 bar),
2
thus adding the cost of CO concentration and pressurization,
2
2
417
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422

ACS Catalysis
Research Article
Scheme 4. CO Capture with Monoethanolamine (a) and
Valinol (b)
of hydrogenation of oxazolidinones. In the search for a system
for oxazolidinone hydrogenation to MeOH, we first examined
the reactivity of various Ru catalysts in hydrogenation of 3-
methyl-2-oxazolidinone as a model substrate (Table 2). This
model reaction was performed in THF as a solvent using pincer
Ru catalysts developed in our group (A, B, and C). Catalyst A is
currently commercially available. An additive of 1 equiv of base,
2
t
BuOK, was required in order to activate the catalyst precursors
A and B and form a dearomatized catalytically active species A′
2
4,25
and B′, respectively, in situ (Chart 1).
Two equivalents of
t
BuOK was used to activate complex C for the in situ
26
generation of C′, as shown in previous studies (Chart 1).
Comparison of catalysts A, B, and C at 0.5 mol % catalyst
loading shows that catalyst B gives the highest yields of MeOH
and 2-(methylamino)ethanol 8 under these conditions (Table
an alternative for monoethanolamine in CO capture processes,
2
22
and it has a lower commercial price compared to valinol.
2
, entries 1−3), 70 and 80%, respectively. Better yields of
When 2-(methylamino)ethanol was heated under 3 bar of CO2
MeOH and 8 can be obtained at similar times using 1 mol %
catalyst loading (Table 2, entries 4 and 5). Using a THF/water
in DMSO in the presence of 15 mol % of Cs CO for 72 h at
2
3
150 °C, the corresponding 3-methyl-2-oxazolidinone product 9
(
2:1) solution leads to very low conversions (Table 2, entries 6
and 7).
Finally, hydrogenation of 3-methyl-2-oxazolidinone was
examined in DMSO as a solvent, as the Cs CO /DMSO
was obtained in ∼64−66% crude yield (Table 1, entry 3). The
other products of the reaction are most likely bicarbonate and
carbamic acid salts of the protonated 2-(methylamino)ethanol,
which were previously reported as byproducts of CO capture
2
3
2
22
system was shown to be optimal for CO capture to selectively
with 2-(methylamino)ethanol. Attempted replacement of
Cs CO3 with less expensive K CO failed to produce
2
form oxazolidinones. While DMSO was not the ideal solvent
for hydrogenation of 3-methyl-2-oxazolidinone at 1 mol %
catalyst loading (Table 2, entry 8), increasing the catalyst
loading to 2 mol % and prolonging reaction time led to efficient
conversion of 3-methyl-2-oxazolidinone to 2-(methylamino)-
ethanol and MeOH in 96 and 78% yields, respectively (Table 2,
2
2
3
satisfactory results (compare entries 1 and 4). The reaction
catalyzed by Cs CO in less polar solvents, such as THF or
2
3
dioxane, also led to low yields of the product, likely due to the
low solubility of Cs CO in these solvents (compare entries 1,
2
3
5
, and 6).
The CO2 capture with 2-(methylamino)ethanol is also
entry 9). Under these conditions, less than 2% of Me S was
2
n
catalyzed by Bu SnO in toluene; however, this catalytic system
formed through DMSO hydrogenation, suggesting that the
catalyst was selective toward hydrogenation of the oxazolidi-
none substrate in the presence of DMSO.
2
typically required higher temperature to obtain comparable
yields of 3-methyl-2-oxazolidinone (entry 7). In addition,
n
Bu SnO is not compatible with the Ru pincer catalyst that is
Similarly, hydrogenation of 4-isopropyl-2-oxazolidinone in
2
t
used for hydrogenation of oxazolidinone products to MeOH
DMSO in the presence of 2 mol % of B and 2 mol % of BuOK
(
see next sections).
These results indicate that using a simple catalyst, Cs CO ,
led to the selective formation of valinol (>99%) and MeOH
(83% yield) in DMSO as a solvent (Scheme 5a). For
comparison, hydrogenation of 7 in DMSO under similar
conditions in the presence of the commercially available Ru
2
3
selective CO capture at 1−3 bar of CO can be achieved with
valinol and 2-(methylamino)ethanol in DMSO to give
oxazolidinones as the major products.
2
2
t
catalyst, RuMACHO (2 mol %) and BuOK (2 mol %), leads to
Hydrogenation of Oxazolidinones: Catalyst and
Reaction Condition Screening. Hydrogenation of noncyclic
carbamic esters using a Ru pincer catalyst has been previously
reported by our group; however, cyclic substrates were not
an unselective reaction and the formation of only 10% of
MeOH at 41% conversion of 7 (see Supporting Information for
more detail).
Overall, these results demonstrate that catalyst B is the most
efficient catalyst for oxazolidinone hydrogenation to form
23
examined in this study. We are unaware of literature reports
a
Table 1. CO Capture with 2-(Methylamino)ethanol
2
b
entry
catalyst (mol %)
Cs CO (10)
solvent
CO pressure (bar)
T (°C)
t (h)
conversion (%)
yield of 9 (%)
2
1
2
3
4
5
6
7
DMSO
DMSO
DMSO
DMSO
THF
1
3
3
1
1
1
1
150
160
150
150
150
150
160
48
48
53
48
48
48
54
95
99
50
70
2
3
Cs CO (15)
2
3
Cs CO (15)
99
66
2
3
K CO (10)
95
<30
2
3
c
Cs CO (10)
<10
<10
98
nd
2
3
c
Cs CO (10)
dioxane
toluene
nd
2
3
n
Bu SnO (10)
2
64
a
Typical reaction conditions: substrate 8 (1 mmol), catalyst, and solvent were heated in a Fischer−Porter tube under 1−8 bar of CO pressure.
2
b
c
Yields were determined by NMR using pyridine or DMSO as an internal standard. Not detected.
2
418
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422

ACS Catalysis
Research Article
a
Table 2. Ru-Catalyzed Hydrogenation of 3-Methyl-2-oxazolidinone
b
b
entry
1
catalyst (mol %)
solvent
THF
t (h)
conversion (%)
57
yield of 8 (%)
yield of MeOH (%)
51
A (0.5)
23
57
80
64
95
100
7
t
BuOK (0.5)
2
3
4
5
6
7
8
9
B (0.5)
THF
THF
THF
THF
19
19
19
19
19
21
21
48
80
64
70
56
84
92
7
t
BuOK (0.5)
C (0.5)
t
BuOK (1)
A (1)
>99
>99
8
t
BuOK (1)
B (1)
t
BuOK (1)
c
c
A (1)
THF/H O
2
t
BuOK (1)
B (1)
THF/H O
4
<1
36
96
<1
32
78
2
t
BuOK (1)
B (1)
DMSO
DMSO
64
t
BuOK (1)
B (2)
>92
t
BuOK (2)
a
t
Typical reaction conditions: substrate 9, 0.5−2 mol % of Ru catalyst, and 0.5−2 mol % of BuOK were heated at 135 °C under 60 bar of H in a
2
b
stainless steel autoclave; the reaction mixtures were analyzed by NMR and GC-MS. Yields were determined by NMR integration versus internal
standard. 2:1 v/v THF/H O ratio.
c
2
Scheme 5. Hydrogenation of 4-Isopropyl-2-oxazolidinone
a) and 2-Oxazolidinone (b) Catalyzed by B
THF as a solvent. However, under these conditions, no MeOH
was formed, and the major product, formed in 97% yield, was
identified as N-(2-hydroxyethyl)-N-methylformamide (OHC)-
N(Me)(CH CH OH) by ESI and GC-MS (Scheme 6). The
(
2
2
Scheme 6. Attempted Direct Hydrogenation of CO in the
2
Presence of 2-(Methylamino)ethanol, Catalyst B, and
Cs CO
2
3
selectively the corresponding aminoalcohol and methanol in
high yields. This catalytic reactivity is general and is applied to
other substrates, as well, including hydrogenation of an
unsubstituted 2-oxazolidinone, although the latter required
higher catalyst loading (5 mol %) to achieve good yields and
conversions (Scheme 5b).
The high yields of MeOH obtained by hydrogenation of 3-
methyl-2-oxazolidinone and 4-isopropyl-2-oxazolidinone in
DMSO as a solvent imply that B can be a suitable catalyst
formation of the formamide was also observed in the absence of
Cs CO and likely results from Ru-catalyzed CO hydro-
genation in the presence of a secondary amine. Similarly,
screening of other reaction conditions for hydrogenation of
2
3
2
2
7
CO
directly in the presence of 2-(methylamino)ethanol 8 or
2
t
valinol 6 in a solution containing B/ BuOK and Cs
CO
SnO) failed to produce MeOH (see Supporting
Information for more detail).
(or
2
3
n
Bu
2
for developing a combined process for CO capture at low
2
pressure coupled with hydrogenation to MeOH, as shown
below.
These results suggest that catalyst B cannot effectively
hydrogenate formamides (or carbamates) in the presence of
Combining CO Capture and Hydrogenation To Form
CO gas. Complex B could also be modified or deactivated via
2
2
MeOH. Based on the results above, we decided to explore a
the reactivity with CO studied previously by our group and by
Sanford.
2
28,29
combined one-step process for CO hydrogenation to MeOH
2
at low CO pressure, without isolation of the intermediate.
We then proposed that the CO capture and Ru-catalyzed
hydrogenation steps should be performed consecutively to
2
2
First, we attempted direct CO2 hydrogenation in the
presence of 2-(methylamino)ethanol 8 under pressurized
CO (20 bar) and H (60 bar) in the presence of Cs CO
3
avoid exposure of catalyst B to CO gas. However, since both
2
CO2 capture and oxazolidinone hydrogenation can be
performed in DMSO, both steps can be carried out in the
2
2
2
t
(
10 mol %), B (1 mol %), and BuOK (1 mol %) in DMSO or
2
419
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422

ACS Catalysis
Research Article
same solution, without isolation of oxazolidinone. This would
respectively, after heating at 135 °C for 72 h (Table 3, entry 1).
However, when less amounts of BuOK were used (entries 2
t
provide a simple protocol for captured CO hydrogenation to
2
MeOH that avoids the intermediate isolation or purification
steps.
and 3), the yield of MeOH decreased.
This modified protocol was first examined using 2-
Table 3. CO Capture/Hydrogenation to MeOH Using
Valinol
2
(
methylamino)ethanol 8 as a CO capture agent (Scheme 7).
a
2
Scheme 7. CO Capture/Hydrogenation to MeOH Using 2-
2
(Methylamino)ethanol
b
b
b
HCOO−
b
catalyst
(mol %)
yield of
yield of
yield of
entry
1
7 (%)
MeOH (%)
6 (%)
(mmol)
B (2.5)
9
37
62
0.193
0.260
0.141
0.057
t
BuOK (25)
2
3
B (2.5)
39
67
21
6
34
∼12
74
t
First, the CO capture step was performed as described above,
BuOK (15)
2
using 15 mol % of Cs CO in DMSO and 3 bar of CO . After
B (2.5)
2
3
2
t
being heated for 72 h at 150 °C, the resulting solution
containing 66% of 3-methyl-2-oxazolidinone 9 was evacuated at
BuOK (6)
c
e
4
B (2.5)
nd
53
45
44
t
room temperature to remove excess CO gas and was then
BuOK (25)
2
t
d
e
combined with catalyst B (5 mol %) and BuOK (25 mol %).
5
B (2.5)
36
6
63
nd
Subsequent hydrogenation under 60 bar of H at 135 °C
t
2
BuOK (25)
produced MeOH and 2-(methylamino)ethanol 8 in 29 and
c
6
B (2.5)
63
0.089
5
2% yields, respectively, after 69 h. Formate salt was also
t
BuOK (10)
present as a major byproduct (24% yield based on 8) that likely
a
Typical reaction conditions: 1 mmol of valinol and 0.1 mmol of
results from a Ru-catalyzed hydrogenation of remaining CO or
2
Cs CO in DMSO were heated at 150 °C for 24 h; the reaction
30−33
2
3
hydrocarbonate salts.
mixture was then degassed under vacuum at room temperature for 20
t
t
A large excess of BuOK (25 mol %) relative to the catalyst is
min; B and BuOK were added, and the reaction mixture was heated
b
required in this reaction to obtain significant conversion to
under H (60 bar) at 135 °C for 72 h. Yields after the hydrogenation
2
t
c
d
MeOH. For comparison, when 1 equiv of BuOK was used,
step. Reaction mixture was filtered before hydrogenation. Reaction
mixture was diluted with toluene and filtered before hydrogenation.
only a trace amount of MeOH was formed, while starting
materials remained mostly unreacted. Excess base is most likely
e
Not detected.
needed to neutralize acidic byproducts of CO capture (e.g.,
2
hydrocarbonate and formate salts of a protonated 2-
(
methylamino)ethanol), which may react with a dearomatized
Because formate salt was still present among the reaction
products, this could indicate that inorganic impurities such as
cesium or potassium hydrocarbonate salts could still be present
in the reaction mixture and undergo further hydrogenation in
catalyst B′, leading to its deactivation.
An analogous protocol using a lower catalyst loading, 2.5 mol
t
%
of B and 20 mol % of BuOK, produced MeOH and 2-
30−33
(
4
methylamino)ethanol in 21 and 38% yields, respectively, after
the presence of a Ru catalyst upon heating.
possible inorganic contaminants, the reaction mixture after CO
2
To remove
8 h. A control experiment showed that when Cs CO alone
2
3
t
was heated in DMSO under 60 bar of H in the presence of B
capture step was combined with B (2.5 mol %) and BuOK (25
mol %) and then filtered through a Celite plug at room
temperature, and the resulting clear solution was subjected to
2
t
(
10 mol %) and BuOK (10 mol %), no methanol formation
was observed, indicating that Cs CO does not act as a source
2
3
of MeOH. Thus, these initial results demonstrate that MeOH
can be obtained in a total yield of 29% in a simple reaction
sequence and in the same solution, without intermediate
product isolation or purification steps.
typical hydrogenation conditions (60 bar of H , 135 °C, 72 h).
2
This led to improved yields of MeOH and valinol, which were
obtained in 53 and 74% yields, respectively (Table 3, entry 4).
Accordingly, the fraction of formate salts significantly
decreased. Attempted precipitation of inorganic salts with an
equal volume of toluene from DMSO solution followed by
filtration did not improve the results, and MeOH and valinol
were obtained in 45 and 63% yields, respectively, under
analogous conditions (entry 5).
To further improve this process, this protocol was then
tested using valinol 6, which was shown to capture CO more
2
selectively under only 1 bar of CO . First, CO capture was
2
2
performed as described above, using 10 mol % of Cs CO3
2
under 1 bar of CO in DMSO to produce 4-isopropyl-2-
2
oxazolidinone 7 in >90% yield after heating for 24 h at 150 °C
This improved protocol also allowed us to lower the amount
of BuOK to 10 mol % to obtain MeOH and valinol in
comparable yields, 44 and 63%, respectively, under analogous
conditions (entry 6).
t
(
see above). After removal of CO under vacuum and addition
2
t
of B (2.5 mol %) and BuOK (25 mol %), hydrogenation under
0 bar of H produced MeOH and valinol in 37 and 62% yields,
6
2
2
420
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422

ACS Catalysis
Research Article
t
Overall, these results demonstrate that selective CO capture
bar was charged with the Ru complex (10 μmol), BuOK (10−
2
under mild conditions (1 bar of CO ) can be combined with
20 μmol), oxazolidinone substrate (1 or 2 mmol), and 1.5 mL
2
Ru-catalyzed hydrogenation of in situ formed oxazolidinone
product to MeOH and aminoalcohol in a simple procedure that
does not require isolation or purification of the captured
oxazolidinone product. Although the yields in such a combined
procedure are moderate, our study of the catalytic activity of
complex B in hydrogenation of oxazolidinones (see above)
shows that these results can potentially be improved by
developing a process engineering solution and improving the
catalyst activity. Presently, complex B is the only reported
catalyst for the highly selective hydrogenation of oxazolidinones
to generate MeOH and the corresponding aminoalcohol.
of the solvent. The autoclave was filled with H gas (60 bar)
2
and heated at 135 °C for an indicated period of time. The
reaction mixture was then cooled before releasing H pressure,
2
then 10 or 20 μL of pyridine was added as a standard, and a
sample of the reaction mixture (50−100 μL) was dissolved in
1
0.6 mL of D O or CDCl and analyzed by H NMR and GC-
2
3
MS. Yields were determined by NMR integration versus
pyridine as an internal standard.
The unaccounted mass balance for MeOH formation (for
example, Table 2, entry 9) could be due to partial loss of a
volatile MeOH product in the headspace.
Typical Procedure for Combined CO Capture/Hydro-
2
genation. A Fischer−Porter tube equipped with a magnetic
CONCLUSION
■
stirring bar was charged with Cs CO , valinol, or 2-
2
3
In summary, we have demonstrated a novel approach to CO2
hydrogenation to MeOH in which CO capture at low pressure
with aminoalcohols is coupled with hydrogenation of the
captured product, oxazolidinone, to form MeOH. This
approach was inspired by the CO capture industry which
(
methylamino)alcohol substrate (1 mmol) and 1 mL of
2
DMSO and then filled with CO at 1 bar (for reaction with
2
valinol) or 3 bar (for reaction with 2-(methylaminoethanol).
The reaction mixture was heated at 150 °C for 24 h (for
reaction with valinol) or 53 h (for reaction with 2-
2
utilizes aminoalcohols to capture CO from waste streams.
2
(
methylamino)ethanol). The solution was then cooled and
However, the capture of oxazolidinone product has previously
not been considered a useful precursor for the production of
liquid fuel such as MeOH due to the lack of methods for such
conversion. We have shown here that the Ru pincer complexes
A, B, and C) are active catalysts for the unprecedented
hydrogenation of oxazolidinones, which are often formed as
stirred under vacuum for 20 min at room temperature to
remove CO . A solution of complex B (25−50 μmol) and
2
t
BuOK (60−250 μmol) in 2 mL of DMSO was then added.
The reaction mixture was then transferred to a 45 mL autoclave
(
equipped with a magnetic stirring bar, filled with H (60 bar),
2
and heated at 135 °C for 72 h. After being cooled, H was
2
byproducts of CO capture. Moreover, we have shown that
2
released and 20 μL of pyridine was added as a standard. A
sample of the reaction mixture (50−100 μL) was dissolved in
both steps, CO capture to selectively produce oxazolidinones
2
and their subsequent hydrogenation to MeOH, can be
performed in the same reaction mixture using a simple protocol
that avoids intermediate isolation or purification steps. For
0
.6 mL of D O (or CDCl in entry 5, Table 3) and analyzed by
2 3
1
H NMR. The yields were determined versus pyridine as an
internal standard. In an optimized procedure (entries 4−6,
example, using valinol for selective CO capture at 1 bar
2
t
Table 3), the reaction mixture after addition of B and BuOK
catalyzed by Cs CO and hydrogenation of corresponding
2
3
was filtered through a short Celite plug before being transferred
to an autoclave.
oxazolidinone using catalyst B allowed us to obtain MeOH in
up to ∼50% total yield.
The advantage of this approach is that it allows one to utilize
ASSOCIATED CONTENT
the CO capture product directly for MeOH production, thus
■
2
avoiding the energy costs associated with CO regeneration
from capture products and pressurization. This is conceptually
different from other previously reported catalytic processes
where pressurized CO is used to produce MeOH. We hope
that these conceptually new results will stimulate the
development of novel processes for direct utilization of CO2
capture products to produce liquid fuels or other value-added
products.
*
S
Supporting Information
2
The following file is available free of charge on the ACS
Publications website at DOI: 10.1021/acscatal.5b00194.
2
General information, experimental details, and character-
ization data (PDF)
AUTHOR INFORMATION
■
Corresponding Author
E-mail: david.milstein@weizmann.ac.il.
EXPERIMENTAL SECTION
*
■
Typical Procedure for CO Capture. A Fischer−Porter
pressure tube equipped with a magnetic stirring bar was
Present Address
2
†
(
J.R.K.) Okinawa Institute of Science and Technology
charged with Cs CO (0.1−0.15 mmol), valinol, or 2-
2
3
Graduate University, 1919-1 Tancha, Onna-son, Kunigami-
gun, Okinawa, Japan 904-0495.
(
methylamino)ethanol (1 mmol) and a preweighed amount
of DMSO (1 mL). The Fischer−Porter tube was filled with
Notes
CO gas to 1 bar (for valinol) or 3 bar (for 2-(methylamino)-
2
The authors declare no competing financial interest.
ethanol) pressure and heated at 150 °C for 24−72 h. The
solution was cooled, and CO was released. A sample of the
reaction mixture was dissolved in D O or CDCl and analyzed
by NMR. Yields were determined versus DMSO as an internal
standard.
Typical Procedure for Hydrogenation of Oxazolidi-
nones. In a nitrogen-filled glovebox, an oven-dried 45 mL
autoclave with a Teflon insert equipped with a magnetic stirring
2
ACKNOWLEDGMENTS
■
2
3
This research was supported by the European Research Council
under the FP7 framework (ERC No. 246837), by the Israel
Science Foundation, and by the Kimmel Center for Molecular
Design. We thank the Swiss Friends of The Weizmann Institute
of Science for a generous Post Doctoral Fellowship to J.A.G.
2
421
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422

ACS Catalysis
REFERENCES
Research Article
■
(
1) Olah, G. A.; Goeppert, A.; Prakash, G. K. S. Beyond Oil and Gas:
The Methanol Economy; Wiley-VCH: Weinheim, Germany, 2009; p
3
(
34.
2) Goeppert, A.; Czaun, M.; Jones, J.-P.; Surya Prakash, G. K.; Olah,
G. A. Chem. Soc. Rev. 2014, 43, 7995−8048.
(
(
3) Rochelle, G. T. Science 2009, 325, 1652−1654.
4) Vaidya, P. D.; Kenig, E. Y. Chem. Eng. Technol. 2007, 30, 1467−
1
474.
(
5) Conway, W.; Wang, X.; Fernandes, D.; Burns, R.; Lawrance, G.;
Puxty, G.; Maeder, M. J. Phys. Chem. A 2011, 115, 14340−14349.
6) Rochelle, G. T.; Goff, G. S.; Cullinane, J. T.; Freguia, S. LRGCC
(
2
2
(
(
002 Conference Proceedings, Norman, Oklahoma, February 24−27,
002; pp 131−152.
7) Freguia, S.; Rochelle, G. T. AIChE J. 2003, 49, 1676−1686.
8) Rao, A. B.; Rubin, E. S. Environ. Sci. Technol. 2002, 36, 4467−
4
475.
(
9) Shao, R.; Stangeland, A. The Bellona Foundation. “Amines Used
in CO Capture: Health and Environmental Impacts”; Bellona Report,
2
2
009; http://bellona.org/news/uncategorized/2010-01-amines-used-
in-co2-capture-health-and-environmental-impacts.
(
(
10) Davis, J.; Rochelle, G. Energy Procedia 2009, 1, 327−333.
11) Direct CO hydrogenation in the presence of heterogeneous
2
catalysts typically requires temperatures > 200 °C. See ref 2 and
references therein.
(
12) Tominaga, K.; Sasaki, Y.; Kawai, M.; Watanabe, T.; Saito, M. J.
Chem. Soc., Chem. Commun. 1993, 629−631.
13) Tominaga, K.-i.; Sasaki, Y.; Watanabe, T.; Saito, M. Bull. Chem.
Soc. Jpn. 1995, 68, 2837−2842.
14) Huff, C. A.; Sanford, M. S. J. Am. Chem. Soc. 2011, 133, 18122−
8125.
15) Wesselbaum, S.; vom Stein, T.; Klankermayer, J.; Leitner, W.
Angew. Chem., Int. Ed. 2012, 51, 7499−7502.
16) Rezayee, N. M.; Huff, C. A.; Sanford, M. S. J. Am. Chem. Soc.
015, 137, 1028−1031.
17) Yadav, M.; Linehan, J. C.; Karkamkar, A. J.; van der Eide, E.;
(
(
1
(
(
2
(
Heldebrant, D. J. Inorg. Chem. 2014, 53, 9849−9854.
(
(
18) Tominaga, K.-I.; Sasaki, Y. Synlett 2002, 307−309.
19) Tamura, M.; Honda, M.; Noro, K.; Nakagawa, Y.; Tomishige, K.
J. Catal. 2013, 305, 191−203.
20) Juarez, R.; Concepcion, P.; Corma, A.; Garcia, H. Chem.
Commun. 2010, 46, 4181−4183.
21) Foo, S. W.; Takada, Y.; Yamazaki, Y.; Saito, S. Tetrahedron Lett.
013, 54, 4717−4720.
22) Folgueira, I.; Teijido, I.; Garcia-Abuin, A.; Gomez-Diaz, D.;
Rumbo, A. Energy Fuels 2014, 28, 4737−4745.
23) Balaraman, E.; Gunanathan, C.; Zhang, J.; Shimon, L. J. W.;
Milstein, D. Nat. Chem. 2011, 3, 609−614.
24) Zhang, J.; Leitus, G.; Ben-David, Y.; Milstein, D. J. Am. Chem.
Soc. 2005, 127, 10840−10841.
25) Balaraman, E.; Gnanaprakasam, B.; Shimon, L. J. W.; Milstein,
D. J. Am. Chem. Soc. 2010, 132, 16756−16758.
26) Fogler, E.; Garg, J. A.; Hu, P.; Leitus, G.; Shimon, L. J. W.;
Milstein, D. Chem.Eur. J. 2014, 20, 15727−15731.
27) Jessop, P. G.; Ikariya, T.; Noyori, R. Chem. Rev. 1995, 95, 259−
72.
28) Vogt, M.; Gargir, M.; Iron, M. A.; Diskin-Posner, Y.; Ben-David,
Y.; Milstein, D. Chem. - Eur. J. 2012, 18, 9194−9197.
29) Huff, C. A.; Kampf, J. W.; Sanford, M. S. Organometallics 2012,
1, 4643−4645.
30) Joo, F.; Joo, F.; Nadasdi, L.; Elek, J.; Laurenczy, G.; Nadasdi, L.
Chem. Commun. 1999, 971−972.
31) Federsel, C.; Jackstell, R.; Boddien, A.; Laurenczy, G.; Beller, M.
ChemSusChem 2010, 3, 1048−1050.
32) Langer, R.; Diskin-Posner, Y.; Leitus, G.; Shimon, L. J. W.; Ben-
David, Y.; Milstein, D. Angew. Chem., Int. Ed. 2011, 50, 9948−9952.
33) Ziebart, C.; Federsel, C.; Anbarasan, P.; Jackstell, R.; Baumann,
W.; Spannenberg, A.; Beller, M. J. Am. Chem. Soc. 2012, 134, 20701−
0704.
(
(
2
(
(
(
(
(
(
2
(
(
3
(
(
(
(
2
2
422
DOI: 10.1021/acscatal.5b00194
ACS Catal. 2015, 5, 2416−2422