Communications
(>99%) of TMA·HCl was obtained at an increased catalysts
loading of 5 mol% (Table 3, entry 4).
reaction was performed in the presence of methanol and the
absence of CO2 [Scheme 2, Eq. (3)].
Under the optimized conditions, the aqueous/organic bipha-
sic system also allowed for convenient separation of the cata-
lyst and product (Figure 1). Upon removing the reaction mix-
Taken together, these results suggest, in principle, two possi-
ble pathways for the methylation of ammonia and ammonium
chloride with the novel catalytic process (Scheme 3). In, path-
way A, the methylation occurs via formamide intermediates,
Figure 1. The aqueous/organic biphasic system after the reaction.
ture from the high-pressure reactor, the two liquid phases sep-
arated spontaneously with the catalyst partitioning into the or-
ganic dioxane phase with high preference, as judged from the
coloration. TMA·HCl formed was isolated as a colorless solid
after separation/evaporation of the aqueous phase (Figure 1).
This beneficial phase behavior opens the possibility for inte-
grated reaction/separation sequences under recycling of the
organic catalyst phase.
Scheme 3. Possible reaction pathways for the [Ru(triphos)(tmm)]-catalyzed
methylation of ammonia by using CO2/H2 as a C1 synthon.
which are well known to be formed from amines, CO2, and
H2.[6] In pathway B, methanol is formed from CO2 and H2 and
acts as a methylating agent through typical alcohol-amination
mechanisms.[14] As Ru–triphos is known to catalyze the hydro-
genation of amides[15] as well as the hydrogenation of CO2 to
methanol,[7] distinction between the two pathways is currently
not plausible. Given the possible equilibria between the in-
volved species, they may well occur in parallel under the cata-
lytic action of the Ru–triphos system.
To gain insight into the pathway of the selective formation
of TMA starting from the NH3/CO2/H2 feed, we conducted a set
of experiments with different possible intermediates for TMA
formation. Formamides were previously suggested as inter-
mediates for the direct N-methylation of amines.[6b,13] Under
the standard reaction conditions but in the absence of CO2,
the hydrogenation of DMF gave TMA in a yield of 33% in addi-
tion to trace amounts of methanol and 18% of DMA, which
most likely resulted from catalytic decarbonylation of DMF
[Scheme 2, Eq. (1)]. If dimethylamine was used as a substrate
In conclusion, the selective catalytic triple N-methylation of
ammonia and ammonium chloride by using CO2 as a C1 source
and molecular hydrogen as a reducing agent was demonstrat-
ed. The catalytic system comprised the readily available com-
plex [Ru(triphos)(tmm)] [triphos: 1,1,1-tris(diphenylphosphino-
methyl)ethane, tmm: trimethylene methane] as a precatalyst.
For the selective synthesis of trimethylamine (TMA) from am-
monia in organic solvents, a Lewis and/or Brønsted acid as co-
catalyst was required. The conversion of ammonium chloride
occurred very efficiently with practically quantitative yield in
a two-phase aqueous/organic system. No co-catalyst was re-
quired in this case, probably because of the availability of the
inherent proton from the substrate. Intermediate formation of
formamides and/or methanol were identified as possible path-
ways for the construction of the methyl groups from the CO2/
H2 mixture, and further mechanistic studies are underway to
unravel the complex reaction network.
A salient feature of this new catalytic reaction is the use of
ammonia, CO2, and H2 as the only reagents: all three compo-
nents are readily available at ammonia production sites. At
present, the hydrogen for ammonia production is formed from
fossil feedstocks, which results in concomitant CO2 formation
through the water gas shift equilibrium. Whereas the current
technology to produce methylamines requires an additional
unit operation for methanol production, the catalytic process
described herein allows for direct utilization of the available
gaseous production streams. Of course, the potential impact
on the carbon footprint of the synthesis of TMA or TMA·HCl
Scheme 2. Control reactions to elucidate possible pathways for the [Ru(tri-
phos)(tmm)]-catalyzed methylation of ammonia starting from DMF, DMA, or
NH3.
under CO2/H2 pressure, TMA was obtained in 26% yield, and in
addition, DMF was detected as a byproduct [Scheme 2,
Eq. (2)]. Again, however, methanol was observed in trace
amounts under these conditions. This prompted us to probe
the direct methylation of ammonia with methanol under these
mild conditions. Indeed, TMA was obtained in 53% yield if the
ChemCatChem 2016, 8, 135 – 138
137
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim