Production of hcooh/net3 adducts by co2/h2 incorporation into neat NEt3

Article abstract of DOI:10.1002/anie.200906054

Chemical Equaction Presented Adding HCOOH to NEt₃ gives a biphasic system of amine and an adduct with a molar acid/amine ratio (AAR) of 1.33. In the presence of a suitable catalyst, CO₂/H ₂ (1:1) acts as HCOOH if both amine and 1.33-adduct phases are present. For example, at 4O°C and 120 bar, neat amine, doped with both 1.33 adduct and catalyst, is quantitatively converted into 1.78 adduct, which is distilled as an azeotrope with AAR = 2.35.

Full text of DOI:10.1002/anie.200906054

Angewandte
Chemie
DOI: 10.1002/anie.200906054
Hydrogen Storage
Production of HCOOH/NEt Adducts by CO /H Incorporation into
3
2
2
Neat NEt **
3
Debora Preti, Sergio Squarcialupi, and Giuseppe Fachinetti*
Research efforts are currently devoted to hydrogen storage in
genation in neat NEt , provided the reaction can be driven to
3
[
1]
CO as formic acid derivatives, a thirty-year old idea not yet
completion. As far as we know, only three reports mention
2
I
implemented. A chemical loop is conceivable based on a
formate intermediate which can be produced in a pure and
storable form under moderately high CO and H pressures
hydrogenation reactions in neat NEt : both Rh with a
3
[
7]
[8]
bidentate phosphane ligand and [RuCl(O CMe)(PMe ) ],
2
3 4
which promote the reaction in DMSO/NEt , were found to be
2
2
3
and which can release gases at atmospheric pressure on
demand when in contact with an appropriate catalyst. The
catalytic decomposition of HCOOH itself has received
inactive in neat amine. More encouragingly, a third paper
[9]
reports that [RuCl(O CMe)(PMe ) ] promotes CO hydro-
2
3
4
2
genation in neat NEt , but the reaction was stopped by
3
[
2]
renewed attention. Remarkably, M. Beller and co-workers
freezing when only 15% of NEt had reacted. For better
3
have demonstrated that HCOOH/NEt adducts can also be
understanding and planning of CO /H incorporation into
3
2
2
catalytically decomposed at 408C to H free of CO, albeit
NEt it must be borne in mind that addition of HCOOH to an
excess of amine leads to a biphasic system constituted by
2
3
[
3]
[10]
diluted by CO2. However, a procedure for the production of
either HCOOH or HCOOH/NEt adducts from CO and H
NEt and an adduct with AAR = 1.33, whose composition
3
2
2
3
is still lacking. The formation of HCOOH by CO hydro-
remains unchanged upon addition of HCOOH until free
amine is present (see Supporting Information). The chemical
and physical properties of the amine are completely different
from those of the immiscible adduct. For instance, the
dielectric constant of the amine is 2.4, while that of the 1.33
2
genation is hampered by unfavorable thermodynamics
o
98
ꢀ1
(
DG2 = + 33 kJmol
for the reaction H2(g) + CO2(g)
!
HCOOH ). The HCOOH/NEt adducts are reportedly
(
l)
3
formed under moderate CO /H2 pressures, with a molar
2
acid/amine ratio (AAR) in the range of 1.3–1.8, by employing
adduct is so high that cannot be determined by classical
II
I
[10]
a variety of Ru and Rh homogeneous precursors. High
turnover numbers and frequencies (TOF) have been achieved
methods and it has been evaluated to be 86.6.
The IR
spectra of the two equilibrated phases (amine and the 1.33
adduct) show that the amine contains only traces of undis-
[4]
in supercritical carbon dioxide (scCO ) or organic solvents,
2
ꢀ
1
but the large amounts of scCO or organic solvents that are
sociated HCOOH ( ~n = 1706 cm ), while the 1.33 adduct is
CO
+
2
ꢀ
ꢀ1
needed render these methods unsuitable for industrial
application. Moreover, separation of the adduct from the
reaction medium and of the homogeneous catalyst from the
basically a [NEt H] [HCOO] ( ~n = 1598 cm ) ionic liquid
3 CO
ꢀ
1
with some oligomeric (HCOOH)n ( ~n CO = 1710 cm ; see
Supporting Information). An intriguing hypothesis is that
CO /H incorporation occurs in the separate phase of the 1.33
[
5]
adduct has been achieved only in the case of scCO2 by
employing [RuH (PMe ) ] as homogeneous catalyst. Under
2
2
adduct rather than in NEt itself, and that the amine merely
2
3
4
3
these conditions, removal of gases at the end of the reaction
can be performed without product entrainment, and the
catalyst contaminating the adduct can be deactivated by
converts the richer adduct back to the 1.33 adduct. Under
these conditions, the presence of a separate phase of the 1.33
adduct would be essential to start the reaction, which would
be autocatalytic. This is what we observed: with [RuCl2-
(PMe ) ] (1) as promoter, a long induction period occurred in
exposure to air. The use of a heterogeneous catalyst for CO
2
[
6]
hydrogenation in the presence of NEt3 would make catalyst
removal easy.
3
4
neat NEt , but the incorporation reaction promptly started if
3
We report here the quantitative conversion of NEt to
the 1.33 adduct was present as a separate phase. Therefore, in
order to circumvent the induction period, we invariably
3
pure and storable HCOOH/NEt adducts by incorporation of
3
CO /H into neat NEt , without foreign solvents or additives.
employed NEt in the presence of a separate phase of the 1.33
2
2
3
3
Catalyst removal and recycling apart, pure adducts could
be straightforwardly formed by carrying out CO2 hydro-
adduct, that is, we always added 2 wt% HCOOH to the amine
already charged into the reactor. For a systematic investiga-
tion of the incorporation it was necessary to avoid changes in
the gas-phase composition with progress of the reaction.
Therefore, 1-promoted formation of adducts was investigated
by feeding the reactor with an equimolar CO /H mixture
[
*] Dr. D. Preti, Prof. Dr. G. Fachinetti
Dipartimento di Chimica, Universitꢀ di Pisa, 56126 Pisa (Italy)
Fax: (+39)050-221-9246
2
2
prepared in a separate high-pressure reservoir. The amounts
of gaseous mixture fed into the reactor were determined by
weighing the reactor or the reservoir. The reaction order with
E-mail: funfac@dcci.unipi.it
Dr. S. Squarcialupi
Chimet SpA, 52041 Badia Al Pino (Italy)
respect to CO or H has not yet been investigated.
[**] This research was supported by Chimet SpA, Badia Al Pino, Italy. We
2
2
thank Prof. F. Calderazzo for helpful discussions.
Runs were carried out at 40 and 508C, in a magnetically
driven 320 mL Parr autoclave modified to 190 mL by a home-
made Teflon filling, equipped with an internal temperature
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200906054.
Angew. Chem. Int. Ed. 2010, 49, 2581 –2584
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2581

Communications
sensor, heater, and cooler and charged with 80 mL (58 g,
employing 1 as promoter for the decomposition reaction of
the 1.33 adduct. In the same reactor at 40 and 508C, the
pressure built up by the decomposition reaction leveled off at
the same values as those for the formation reaction (Figure 1).
In the equilibrated mixtures two separate liquid phases are
5
3
68 mmol) of amine “doped” with 21 wt% HCOOH and with
mm (1). The autoclave was pressurized with various weighed
amounts of the CO /H equimolar mixture (CO /H :NEt <
2
2
2
2
3
1
.33); for each run, an indication of the reaction rate at
various stages can be derived from the pressure changes with
time (Figure 1).
present: an upper layer of NEt , and a lower layer of 1.33
3
adduct (see mass balances in the Supporting Information).
The one-to-one correspondence between temperature and
pressure is a prominent feature of the equilibrium between
CO /H (1/1), NEt , and the 1.33 HCOOH/NEt adduct. The
The sigmoidal pressure–time curves, together with the
induction period, confirm that the reaction is autocatalytic
and support that the 1.33 adduct incorporates CO /H .
2
2
2
2
3
3
Remarkably, the pressure leveled off at 41 ꢁ 1 bar at 408C
and at 61 ꢁ 1 bar at 508C regardless of the amount of
equimolar CO /H mixture which had been introduced into
the reactor. Such a one-to-one correspondence between
temperature and equilibrating pressure was confirmed by
Gibbs phase rule (F = CꢀP + 2) establishes a single degree of
freedom of the intensive properties temperature and pressure
for a system with two components (C) and three phases (P).
Indeed, due to the constant CO /H molar ratio of 1/1 and the
2
2
2
2
working equilibrium, only two independent components are
present, while the three phases are a consequence of the
immiscibility between the amine and the 1.33 adduct. In
principle, therefore, it is expected that the 1.33 adduct is
quantitatively formed or decomposed, respectively, at CO /H
2
2
pressures just above or just below the equilibrium value at a
given temperature. However, for complete conversion of
amine to adducts, kinetic reasons, that is, concurrent catalyst
deactivation, compelled us to avoid temperatures higher than
4
08C and to use working pressures considerably higher than
that of equilibrium (41 bar). Figure 2 shows that, under
Figure 2. Moles of incorporated CO /H equimolar mixture per mole
2
2
of NEt at various times, under constant pressures of 80 or 120 bar at
3
40 or 508C, respectively. 190 mL reactor charged with 70.5 mL (51 g,
0
.5 mol) of NEt “doped” with 2 wt% HCOOH and 3 mm 1. *:
3
T=408C, P=120 bar; *: T=408C, P=80 bar; ~: T=508C,
P=120 bar; ~: T=508C, P=80 bar.
1
20 bar pressure at 408C, the incorporation reaction proceeds
beyond AAR = 1.33 up to AAR = 1.78. Both under lower
pressure and at higher temperature the incorporation reac-
tion fades before amine conversion is complete, that is, before
the incorporated number of moles of CO /H per mole of
2
2
NEt reaches a value of 1.33.
3
Figure 1. Time course of pressure during 1-promoted equilibration
Such behavior is due to catalyst deactivation competing
with CO /H incorporation. At a given temperature, increas-
between the 1.33 HCOOH/NEt adduct and NEt under different CO /
3
3
2
2
2
H (1/1) initial pressures. Reactor volume=190 mL; liquid-phase
2
ing pressure favors incorporation but does not affect catalyst
deactivation. Under a given pressure, a higher temperature
accelerates catalyst decomposition more than incorporation,
volume=80 mL; weighed amounts of the introduced gas are given in
the inset. A) T=408C. B) T=508C. ^
: CO /H =23.2 g; &: CO /
2
2
2
H =15.6 g; ^: 1.33 AAR decomposition.
2
2
582
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 2581 –2584

Angewandte
Chemie
presumably due to a larger positive activation enthalpy of the
former.
The incorporation reaction is exothermic, and under
ꢀ
1
favorable conditions for which a TOF of 1600 h is reached
Figure 3), effective cooling is needed to avoid an undesirable
(
rise in temperature and catalyst deactivation.
Figure 3. TOFs for incorporation of equimolar CO /H mixture at
2
2
various P and T with [1]=3 mm. *: T=408C, P=120 bar; *: T=408C,
P=80 bar; ~: T=508C, P=120 bar; ~: T=508C, P=80 bar.
Figure 4. Relationship between equilibrating pressure and AAR>1.33
at 408C. Equilibrating pressures were obtained from incorporation or
decomposition reaction. *: incorporation;
poration, T=408C, P=120 bar.
&
: decomposition; *: incor-
On the basis of these findings, the adducts, rather than the
amine itself, seem to be the active species incorporating CO2/
H . In fact, it was shown above (Figure 1) that 1-promoted
2
incorporation into NEt is autocatalytic and depends on the
To summarize, for the first time a pure formic acid
derivative has been isolated in substantial amounts by
3
amount of 1.33 adduct. On the other hand, the capability of
adducts to incorporate CO /H in the absence of free amine is
employing CO rather than CO as raw material. At 408C
2
2
2
demonstrated by formation of the 1.78 adduct.
under 120 bar CO /H , NEt is quantitatively converted
2
2
3
To evaluate how high an AAR can be reached under
within hours to a 1.78 adduct whose distillation, after catalyst
deactivation, yielded a pure and stable azeotrope with
AAR = 2.35 without CO or H loss. Catalyst removal and
recycling are easily accomplished by means of limited
amounts of cheap reagents. No foreign solvents or additives
are necessary.
reasonable pressures, that is, how much H can be directly
2
stored in the HCOOH/NEt3 adducts, we investigated the
relationship between equilibrating pressure and AAR > 1.33
at 408C. We assumed that the equilibrating pressure is that
which leveled off at the end of either incorporation or
decomposition reactions. The equilibrating pressure increases
exponentially with increasing AAR (Figure 4), and AAR >
.0 seems unattainable in spite of the low temperature of
08C. For instance, decomposition of a richer adduct at 408C
in a homemade rocking stainless steel autoclave was stopped,
for safety reasons, at 300 bar: under this pressure, presumably
lower than that of equilibrium, the residual adduct still had
AAR = 2.3 (see Supporting Information).
2
2
In conclusion, the process reported here could be
economically carried out even on an industrial scale. The
present findings make hydrogen storage in CO2 more
realistic. Under reasonable pressures, direct incorporation
of CO /H into NEt can afford adducts with a maximum
2
4
2
2
3
AAR of 2. However, adduct decomposition gives back H2/
CO (1/1), and the mixture cannot be employed to feed a fuel
2
cell, in which the accumulation of the inert gas would occur.
To circumvent this problem, it has been proposed that the gas
Finally, by distillation at 688C under reduced pressure, the
.78 adduct was quantitatively converted to recyclable amine
[3b]
1
stream flow through a fuel-cell stack. The profitability of
such a strategy has still to be ascertained. On the other hand,
HCOOH/NEt3 adducts are widely employed in organic
and an azeotrope with AAR = 2.35. To avoid the reverse
decomposition reaction of the adduct, it was necessary to
completely deactivate the ruthenium species before distilla-
tion. This was achieved by treating the ruthenium-contami-
[11]
chemistry as selective hydrogen-transfer reagents,
for
which the here-reported production of adducts could be
economically viable.
ꢀ1
nated 1.78 adduct with KCN in an Ru/CN molar ratio of 6/1
soon after venting the gas at room temperature. Quantitative
ruthenium recycling and de-cyanidation were carried out by
treating the colorless distillation residue with alkaline hypo-
chlorite (Caution: alkali must be in excess, otherwise toxic
Received: October 28, 2009
Revised: January 13, 2010
and explosive RuO is formed; see Supporting Information).
Published online: March 12, 2010
4
Angew. Chem. Int. Ed. 2010, 49, 2581 –2584
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2583

Communications
Keywords: amines · carbon dioxide · hydrogen ·
hydrogen storage · hydrogenation
50, 1603 – 1606; d) D. J. Morris, G. J. Clarkson, J. Guy, M. Wills,
Organometallics 2009, 28, 4133 – 4140; e) B. Loges, A. Boddien,
H. Junge, J. R. Noyes, W. Baumann, M. Beller, Chem. Commun.
.
2009, 4185 – 4187.
[
4] P. G. Jessop in The Handbook of Homogenous Hydrogenation,
Vol. 1 (Eds.: J. G. de Vries, C. J. Elsevier), Wiley-VCH, Wein-
heim, 2007, pp. 490 – 499, and references therein.
[
1] R. Williams, R. S. Crandall, A. Bloom, Appl. Phys. Lett. 1978, 33,
81 – 383.
3
[
2] a) S. Fukuzumi, Eur. J. Inorg. Chem. 2008, 1351 – 1362; b) C.
Fellay, P. J. Dyson, G. Laurenczy, Angew. Chem. 2008, 120, 4030 –
[
[
[
[
[
5] P. J. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc.
1
996, 118, 344 – 355.
6] M. W. Farlow, H. Adkins, J. Am. Chem. Soc. 1935, 57, 2222 –
223.
7] W. Leitner, E. Dinjus, F. Gaßner, J. Organomet. Chem. 1994, 475,
57 – 266.
4
032; Angew. Chem. Int. Ed. 2008, 47, 3966 – 3968; c) X. Zhou, Y.
Huang, W. Xing, C. Liu, J. Liao, T. Lu, Chem. Commun. 2008,
540 – 3542; d) S. Enthaler, ChemSusChem 2008, 1, 801 – 804;
2
3
e) F. Joꢀ, ChemSusChem 2008, 1, 805 – 808; f) S. Fukuzumi, T.
Kobayashi, T. Suenobu, ChemSusChem 2008, 1, 827 – 834; g) C.
Fellay, N. Yan, P. J. Dyson, G. Laurenczy, Chem. Eur. J. 2009, 15,
2
8] P. Munshi, A. D. Main, J. C. Linehan, C.-C. Tai, P. G. Jessop, J.
Am. Chem. Soc. 2002, 124, 7963 – 7971.
9] C. A. Thomas, R. J. Bonilla, Y. Huang, P. G. Jessop, Can. J.
Chem. 2001, 79, 719 – 724.
10] P. Huyskens, N. Felix, A. Janssens, F. Van den Broeck, F.
Kapuku, J. Phys. Chem. 1980, 84, 1387 – 1393.
11] S. Gladiali, E. Alberico, Chem. Soc. Rev. 2006, 35, 226 – 236.
3
4
752 – 3760; h) M. Ojeda, E. Iglesia, Angew. Chem. 2009, 121,
894 – 4897; Angew. Chem. Int. Ed. 2009, 48, 4800 – 4803.
[
3] a) B. Loges, A. Boddien, H. Junge, M. Beller, Angew. Chem.
[
[
2008, 120, 4026 – 4029; Angew. Chem. Int. Ed. 2008, 47, 3962 –
3965; b) A. Boddien, B. Loges, H. Junge, M. Beller, ChemSu-
sChem 2008, 1, 751 – 758; c) H. Junge, A. Boddien, F. Capitta, B.
Loges, J. R. Noyes, S. Gladiali, M. Beller, Tetrahedron Lett. 2009,
2
584
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 2581 –2584