Communications
time) and using the [{RuCl (benzene)} ] homogeneous catalyst,
[
a]
2
2
Table 1. Catalytic hydrogenation of bicarbonates and carbonates.
a 35% formate yield with a TON of 807 was achieved from hy-
drogenating 24 mmol NaHCO in a solution of 25 mL H O and
[
b]
[c]
Entry
Reagent
Catalyst
Reaction conditions
Yield
[%]
TON
3
2
p
H2 [MPa]
t
À1 [5a]
5 mL THF (reagent concentration <1 molL ).
Our results
[h]
suggest that the hydrogen storage process based on the re-
duction of ammonium bicarbonate over the Pd/AC heteroge-
neous catalyst is more efficient than the homogeneous coun-
terpart that Beller et al. reported.
1
2
3
4
5
6
7
8
9
NaHCO
Na CO
KHCO
CO
NH HCO
(NH CO
3
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Pd/AC
Ru/AC
Rh/AC
Pt/AC
Ni/AC
2.75
2.75
2.75
2.75
2.75
2.75
2.75
2.75
0.69
1.38
4.14
5.52
5.52
2.75
2.75
2.75
2.75
2.75
2.75
2.75
1
1
1
1
1
1
6
15
1
1
1
1
2
1
1
1
1
1
1
1
28.6
0.05
30.8
0.07
42.4
15.1
84.9
95.6
16.9
31.3
53.1
59.6
90.4
0.2
0.2
0
0
8.9
0.6
2.9
527
<1
567
<1
782
278
1571
1769
312
579
982
1103
1672
3
2
3
3
K
2
3
The hydrogenation of ammonium bicarbonate with different
supported metal catalysts was also investigated. Transition
metals supported carbon such as Pd/AC, Ru/AC, Rh/AC, Pt/AC,
and Ni/AC have been proven to be active towards hydrogena-
tion or hydrogenolysis reactions. But due to the different prop-
erties of the metals themselves, they often show different ac-
4
3
4
)
2
3
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
4
4
4
4
4
4
4
4
4
4
4
4
4
4
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
3
3
3
3
3
3
3
3
3
3
3
3
3
3
10
1
1
12
13
14
15
16
17
18
19
20
[13]
tivities in response to a specific reaction. In the hydrogena-
tion of NH HCO , we found that only Pd catalysts showed
4
3
3
0
0
278
20
a clear catalytic activity, whereas other transition metals such
as Ru, Rh, Pt, and Ni were inactive under the test conditions
(
Table 1, entries 8 and 14–17). The activated carbon support
Pd/Al
Pd/CaCO
Pd/BaSO
2
O
3
was superior to other supports including Al O , CaCO , and
2
3
3
3
4
212
BaSO (Table 1, entries 18–20). Because the hydrogenation of
4
aqueous NH HCO is a multiphasic reaction (gas/liquid/solid),
4
3
2
[a] Reaction conditions: 20 mL distilled H O, 20 mmol reagent salt, 0.1 g
catalyst, 208C. [b] All catalyst loadings were 5% metal, Ni/AC sample was
the diffusion of the reactant, H , could be rate limiting due to
2
prepared by impregnation method (nickel nitrate as the precursor).
its low solubility in water. Activated carbon is commonly a hy-
[c] TON is turnover number and it was calculated by the formula: total
[14]
drophobic support and can store H2. H could be locally en-
2
moles of formate formed/(total moles of Pd adatoms ꢂD%), where D% is
the dispersion of metal atoms on the support surface. The D% values are
calculated from the CO chemisorption (Table S1).
riched in the carbon channels or on the surface of the carbon
support from H spillover from the Pd metal. Localized higher
2
H concentrations on the surface of the Pd/AC catalyst may
2
thus be attributed to the higher formate yields. On the other
hand, the dispersion of Pd nanoparticles (NPs) on AC was
higher than that on other supports such as Al O , CaCO , and
We first compared various bicarbonate and carbonate salts
2
3
3
+
+
+
with different cations, Na , K , and NH , for hydrogenation.
BaSO (Table S1) or other supported metal catalysts (Figure S5),
4
4
As shown in Table 1, over Pd nanocatalyst formates could be
easily produced from the hydrogenation of bicarbonates with
all three different cations but not from carbonates [except for
and a higher dispersion implies more active Pd sites that favor
hydrogenation.
In addition to the catalyst materials (metals and supports),
(
NH ) CO ] at room temperature. It was much more difficult to
process conditions such as initial H pressure and reaction tem-
4
2
3
2
hydrogenate carbonate salts than bicarbonate ones as the pro-
tonation of carbonate ions was considered to be the rate-limit-
ing step in aqueous solutions, especially at low tempera-
perature are also the key influencing factors on the hydrogena-
tion of ammonium bicarbonate. Stalder et al. found that a 1:1
hydrogenation reaction equilibrium ratio of formate/bicarbon-
[
12]
tures. The equilibrium between bicarbonate and carbonate
ions highly depends on the pH value of the solution. On in-
creasing the pH value, the equilibrium shifted from bicarbon-
ate to carbonate and thus the formate yield decreased (Fig-
ure S3). In a typical hydrogenation reaction system, 1m
NH HCO aqueous solution (20 mmol NH HCO in 20 mL H O
ate was obtained when purging the NaHCO solution with
3
[15]
0.1 MPa H gas for a long time (40–90 h). In our reaction at
2
a higher H gas pressure of 2.75 MPa, the equilibrium ratio of
2
HCO NH to NH HCO was shifted significantly to approximate-
2
4
4
3
ly 95:5 in 15 h (Table 1, entry 8). On the other hand, on increas-
ing the reaction temperature from 20 to 808C, the hydrogena-
tion rate was increased but the equilibrium yield of formates
was decreased from ~95% to ~50%, as shown in Figure 1.
Higher reaction temperatures favor the dehydrogenation reac-
tion and shift the equilibrium to hydrogen evolution, which is
4
3
4
3
2
and the pH value was 7.7) using the Pd/AC nanocatalyst (AC
stands for activated carbon support, 5 wt% Pd), a high yield of
ammonium formate, ~59.6%, with a turnover number (TON) of
1103 was achieved after reacting for 1 h when the initial H2
[4d,9,10]
pressure was 5.5 MPa. By extending the reaction time to 2 h,
a 90.4% formate yield with a TON of 1672 was obtained. Fur-
thermore, higher concentrations of NH HCO also increased
in agreement with previous studies.
Our results show
that the equilibrium position of the hydrogen storage and evo-
lution process depends on both the reaction temperature and
the hydrogen pressure in the heterogeneously catalytic
system. Note that the highest formate yield was achieved at
room temperature, implying that our hydrogenation reaction
4
3
the formate yields, for example, a high turnover frequency
À1
(
TOF) of 1221 h was achieved using a 2.5m NH HCO solution
4
3
and 2.75 MPa initial H pressure, as seen in Figure S4. In com-
2
[7]
parison, Beller and co-workers reported that, under similar re-
system does not need additional external energy for heating.
action conditions (5 MPa initial H pressure and a 2 h reaction
However, in practice, a higher H pressure might be needed to
2
2
&
ChemSusChem 0000, 00, 0 – 0
2
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!