Sustainable hydrogenation of aliphatic acyclic primary amides to primary amines with recyclable heterogeneous ruthenium-tungsten catalysts

Article abstract of DOI:10.1039/c9gc01310e

The hydrogenation of amides is a straightforward method to produce (possibly bio-based) amines. However current amide hydrogenation catalysts have only been validated in a rather limited range of toxic solvents and the hydrogenation of aliphatic (acyclic) primary amides has rarely been investigated. Here, we report the use of a new and relatively cheap ruthenium-tungsten bimetallic catalyst in the green and benign solvent cyclopentyl methyl ether (CPME). Besides the effect of the Lewis acid promotor, NH₃ partial pressure is identified as the key parameter leading to high primary amine yields. In our model reaction with hexanamide, yields of up to 83% hexylamine could be achieved. Beside the NH₃ partial pressure, we investigated the effect of the catalyst support, PGM-Lewis acid ratio, H₂ pressure, temperature, solvent tolerance and product stability. Finally, the catalyst was characterized and proven to be very stable and highly suitable for the hydrogenation of a broad range of amides.

Full text of DOI:10.1039/c9gc01310e

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: R. Coeck, S.
Berden and D. E. De Vos, Green Chem., 2019, DOI: 10.1039/C9GC01310E.
Volume 20
Number
May 2018
Pages 1919-2160
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
9
7
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Page 1 of 16
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01310E
Sustainable hydrogenation of Aliphatic Acyclic Primary Amides to
Primary Amines with Recyclable Heterogeneous Ruthenium-
Tungsten Catalysts
Robin Coeck, Sarah Berden and Dirk E. De Vos, cMACS, KU Leuven, Leuven,
Belgium
The hydrogenation of amides is a straightforward method to produce (possibly bio-based) amines. However
current amide hydrogenation catalysts have only been validated in a rather limited range of toxic solvents and the
hydrogenation of aliphatic (acyclic) primary amides has been seldomly investigated. Here, we report the use of a
new and relatively cheap ruthenium-tungsten bimetallic catalyst in the green and benign solvent cyclopentyl
methyl ether (CPME). Besides the effect of the Lewis acid promotor, NH
parameter leading to high primary amine yields. In our model reaction with hexanamide, yields of up to 83%
hexylamine could be achieved. Beside the NH partial pressure, we investigated the effect of the catalyst support,
PGM-Lewis acid ratio, H pressure, temperature, solvent tolerance and product stability. Finally, the catalyst was
3
partial pressure is identified as the key
3
2
characterized and proven to be very stable and highly suitable for the hydrogenation of a broad range of amides.
Introduction
Catalytic hydrogenation of amides is a green and sustainable method for the production of
valuable amines. With an annual aliphatic amine consumption of more than 2 Mton, amines
are widely used for the synthesis of e.g. pharmaceuticals, agrochemicals, dyes, polymers,
fabric softeners etc.1-2 However, the hydrogenation of amides is a challenging reaction,
because of their high stability.3
Traditionally, amide reduction is achieved with either copper chromite-based hydrogenation
catalysts or by applying an excess of hydride reagents. Both methods have severe
shortcomings and lack sustainability. The requirement of stoichiometric amounts of hydride
reagents such as LiAlH or NaBH results in large quantities of inorganic chemical waste and
4
4
an extensive product workup, whereas copper chromite-based catalysts require extreme
reaction conditions (e.g. >200 bar H , 250-350°C) and high catalyst loadings, require specific
2
disposal and generally have a low selectivity for the primary amine.4-6
Over the last few decades, progress has been made with the development of heterogeneous
bimetallic catalysts operating under mild reaction conditions. The first such catalyst was
reported in 1988 in a BP patent, which claimed a Pd/Re/high surface area graphite/zeolite 4A
catalyst for the reduction of amides at 130 bar H and 200°C. Currently, a variety of bimetallic
2
catalysts have been reported, containing a platinum group metal (Pt, Pd, Rh or Ru) promoted
by a group 5, 6 or 7 metal oxide (V, Mo or Re); these catalysts are active within a operational
window of 70-200°C and 30-100 bar H2.3,7-23,43 Although several amides have successfully
been hydrogenated to the corresponding amines with high yields and/or selectivity (at least
70% yield), several issues still need to be addressed. First, these catalysts have only been
validated in a surprisingly narrow range of toxic solvents e.g. 1,2-dimethoxyethane (DME),
hexane and 1,4-dioxane. A smart solvent choice however, can drastically reduce pollution,
contribute to a better air quality and reduce the overall environmental footprint of the

Green Chemistry
Page 2 of 16
process.24 Secondly, the hydrogenation of aliphatic, acyclic primary amides has been rarely
View Article Online
DOI: 10.1039/C9GC01310E
investigated with these new bimetallic catalysts; most reports use cyclic or more activated
amides as reactants, such as ε-caprolactam or N-acetylmorpholine. In as far as aliphatic
acyclic primary amides are considered, they are usually converted to the alcohol or
secondary amine, or must be hydrogenated under extreme reaction conditions (e.g. in liquid
6
ammonia, at high temperatures etc.) to “selectively” obtain the primary amine. Nevertheless,
a mild catalytic system able to selectively reduce aliphatic acyclic primary amides, could be a
powerful tool to produce even bio-based primary amines. As of today, several bio-based
amines, e.g. 1,10-diaminodecane and fatty amines, are synthesized from the corresponding
carboxylic acids in a multistep process: first, the carboxylic acid reacts with ammonia at high
temperature; next, the resulting amide is dehydrated over an alumina or zinc oxide catalyst at
280-400°C, and finally, the nitrile is hydrogenated to the amine at a temperature of 80-140°C
with 10-40 bar H , over nickel This energy-intensive
a
catalyst.2,25-29
2
dehydration-hydrogenation process could be significantly shortened by a direct aliphatic
amide reduction, requiring less energy. Additionally, we expect that in the near future this
pathway will receive an increase in interest, since more and more amines may be produced
from bio-based platform molecules, e.g. butyric acid, propionic acid, adipic acid etc.30-31
In this work, we report on RuWO /MgAl O as a new and recyclable catalyst for the selective
x
2
4
hydrogenation of aliphatic primary amides in cyclopentyl methyl ether (CPME) and under
mild hydrogenation conditions, i.e. with limited addition of NH and H . To our knowledge, this
3
2
is the first report of an amide hydrogenation process in CPME as a benign and
environmentally friendly reaction solvent.32-33
Experimental
Supported bimetallic M-M’O catalysts (where M = Ru, Rh or Pt, M’ = V, Mo or W and support
x
=
SiO , TiO or MgAl O spinel) were prepared with 4 wt% M and a M-M’ ratio of 1-8 by
2 2 2 4
impregnation. First, in a typical synthesis of RuWO /MgAl O , RuCl .xH O (0.52 mmol),
x
2
4
3
2
(
NH ) H (W O ) (0.065 mmol of tungsten), MgAl O (1.25 g) and water (12 mL) are mixed
4 10 2 2 7 6 2 4
together and stirred at ambient temperature to evaporate water. Secondly, when all the water
is removed, the pre-catalyst is dried further overnight in an oven at 60°C. Finally, the material
-1
-1
is granulated (250-500 µm) and reduced at 450°C (4°C min , 100 mL min H , for 4 h) in a
2
quartz U-tube. Catalysts with Rh or Pt are first oxidized at 350°C and reduced afterwards.
2+
Initially RuWO /SiO -Ca catalysts were prepared by first stirring RuWO /SiO (0.1 g) in a
x
2
x
2
solution of Ca(OH) in water (0.005 M, 10 mL) overnight. Next the catalyst was filtered and
2
dried in an oven at 150°C for 4 h. However, this lowered the catalytic activity due to tungstate
leaching. This method was improved by depositing 0.025 mmol of Ca2+ (or 0.05 mmol Na+,

Page 3 of 16
Green Chemistry
+
+
View Article Online
DOI: 10.1039/C9GC01310E
K or Cs ) on 1 g of RuWO /SiO via incipient wetness impregnation and drying the material
x
2
at 150°C for 4 h. PtVO /HAP was synthesized as described by Mitsudome (2017).
x
With the prepared catalysts, reactions were performed in a 25 mL high pressure Parr batch
reactor. In a typical reaction, the reactor is filled with an amide (e.g. hexanamide),
RuWO /MgAl O (4 wt% Ru, Ru-W = 1-8, 5 mol% Ru), undecane (internal standard, 20 µL)
x
2
4
and cyclopentyl methyl ether (CPME, 10 mL). Then, the reactor is sealed, purged (3x N ; 3x
2
H ) and successively pressurized with the desired amounts of NH and H (typically 6 bar
2
3
2
NH and 50 bar H ). After an appropriate reaction time at the desired temperature (200°C,
3
2
stirred at 830 rpm), the reactor is cooled down in an ice bath. The pressure is then released
after which the reactor can be opened. The reaction mixture is transferred to a glass reaction
vial (11mL) which is then sealed and centrifuged. The mixture is then analyzed via GC,
GC-MS and NMR.
Results and discussion
At the onset of our research, several Ru-based catalysts were modified with different molar
ratios of cheap group 5 to 7 metal oxides (Mo, W and V). The solids were screened in the
hydrogenation of hexanamide (1) to hexylamine (3), allowing to determine the best metal
combination and composition for the reduction of aliphatic acyclic primary amides. These
initial experiments were performed at 180°C with 40 bar H , 0.5 mol% Ru in DME for 16 h
2
(Figure 1). For all metal oxides, there appears to be a clear optimum at a Ru/Metal oxide
ratio of 8. RuWO /SiO (Ru/W = 8) was proven to be the best catalyst, yielding 38%
x
2
hexylamine (3) at 69% conversion; alternative compositions are RuMoO /SiO and
x
2
RuVO /SiO respectively, with a yield of 35% and 30% hexylamine (3) at 67% and 70%
x
2
conversion. All these catalysts are superior when compared to simple Ru-catalysts, such as
Ru/SiO . Beside the formation of our desired product, hexanol (2) and dihexylamine (4) were
2
formed as well. Especially secondary amine formation is a predominant side reaction.
Interestingly, unsupported Ru-Mo nanoparticles, as already described by Beamson et al.,
allowed to obtain 77% butylamine yield (and 22% butanol) at 160°C, with 100 bar H , in
2
DME. Moreover, further addition of any metal oxide to a catalyst with the same ratio of 8-1,
lowered the catalytic activity. This drop in activity is presumably caused by a coverage of the
Ru particles, lowering the availability of H* species for the reduction of hexanamide. We also
noticed that the Lewis acid metal oxides must be located on (or in contact with) the Ru
particles, since catalysts prepared by consecutive deposition of first W, followed by Ru, did
not show a superior catalytic activity. In our further experiments, we choose to work with
RuWO /SiO , since this catalyst performed best.
x
2

Green Chemistry
Page 4 of 16
View Article Online
DOI: 10.1039/C9GC01310E
Figure 1 – Variation of the catalysts’ metal combination and composition. Reaction conditions: hexanamide (1 mmol), 180°C,
0 bar H ; 0.5 mol% Ru (4 wt% Ru on catalyst, 5 wt% Ru for Ru/C), undecane (20 µL), DME (10 mL), 16 h.
4
2
Next, the solvent tolerance of the RuWO -catalyst was investigated. As mentioned, the
x
commonly used DME is a very toxic compound. Furthermore, it easily forms explosive
peroxides which generally makes it unsafe to work with, especially in contact with H2.34
A
benign, preferably bio-based reaction solvent would be far more desirable. Therefore, the
hydrogenation activity of RuWO /SiO was tested in several solvents, selected from green
x
2
solvent guides (Figure 2).35 The reduction of hexanamide (1) could be successfully achieved
in any ether solvent, but not in an alcohol. We suggest that the amide-catalyst bond is too
weak to sufficiently overcome the catalyst’s interaction with the hydroxyl group of an alcohol
solvent. Ethers interact a lot less with bimetallic catalysts and are therefore excellent solvents
for amide hydrogenation. It can be noticed that the hydrogenation rate increases with
decreasing solvent polarity (~dielectic constant, ε ): DME (ε = 7.2) < MeTHF(ε = 7) < CPME
r
r
r
(
ε = 4.76) < TAME (ε = 2.6). However, the rate of secondary amine formation also increases
r
r
in the same order (Supporting information, Error! Reference source not found.). CPME
was selected as the best reaction solvent for a number of reasons; first, there appears to be
a good balance between hydrogenation activity and secondary amine formation. Secondly,
the low heat of vaporization (CPME: 289.5 kJ/kg vs MeTHF: 364.4 kJ/kg) and extremely low
solubility of water in CPME (0.3 g water/100 g CPME) ensure an easy and energy-saving
workup when applied in a real industrial process. Thirdly, CPME is very stable and resistant
36
to peroxide formation, which makes it safe to work with. After a typical reaction in CPME, a
concentration of less than 0.001M cyclopentane was observed. These traces of
cyclopentane do not interfere with the reaction, thus the solvent can be reused for many

Page 5 of 16
Green Chemistry
View Article Online
cycles. The other cleavage product, methanol, was not detected. Possibly, methanol is
DOI: 10.1039/C9GC01310E
transformed to volatile consecutive products e.g. CH or to methylamine, which migrate to
4
the gas phase.43
Figure 2 – Variation of the reaction solvent for the hydrogenation of hexanamide. Reaction conditions: hexanamide (1 mmol),
2 x 2
180°C, 40 bar H , 1 mol% Ru (RuWO /SiO ), undecane (20 µL), solvent (10 mL), 16 h.
After selecting CPME as a reaction solvent, a broader catalyst screening was performed
Figure 3 & Error! Reference source not found., Error! Reference source not found.).
(
Commercial PGM on carbon supports performed very poorly, except for Ru/C. Such “single
metal catalysts” can be further improved by depositing small amounts of tungsten on the
surface; not only does this improve the catalytic activity, also the selectivity for C-O cleavage
(amine formation) over C-N cleavage (alcohol formation) increases (Figure 3 & Error!
Reference source not found.). However, the formation of secondary amines cannot be
suppressed by tungsten deposition. Nonetheless, since amine condensation requires a
combination of metal and acid catalysis, one may improve the selectivity for the primary
amine by lowering the amount of Brønsted acid sites. This could either be achieved by
wielding more basic (or at least less acidic) supports, e.g. rutile, or by stirring the slightly
acidic RuWO on fumed silica in a basic solution of e.g. Ca(OH) , followed by a drying step
x
2
(
Error! Reference source not found.).37-38 Regarding this second option, despite the
catalyst’s high stability, one must be careful not to leach out any tungsten as tungstate, since
2+
this will lower the catalytic activity. In this catalyst screening, RuWO /SiO -Ca was the
x
2
catalyst with the highest selectivity for the primary amine (63% selectivity, 53% conversion);
the reaction proceeds without hydrolysis of hexanamide (1) (Figure 3 and Error! Reference
2+
source not found.). More remarkably, our RuWO /SiO -Ca
performs better than
x
2
PtVO /Hydroxyapatite (Figure 3), described by Mitsudome et al. in 2017, an extremely active
x

Green Chemistry
Page 6 of 16
View Article Online
catalyst able to reduce a variety of amides at reaction conditions as low as 70°C and 1 bar
DOI: 10.1039/C9GC01310E
7
H .
2
Figure 3 – Variation of the amide hydrogenation catalyst. Reaction conditions: hexanamide (1 mmol), 180°C, 50 bar H
NH ; 0.5 mol% Ru (or same weight other PGM), undecane (20 µL), CPME (10 mL), 16 h.
2
; 0.5 bar
3
To further investigate the amide hydrogenation and unravel the mechanisms behind it,
2+
hydrogenation experiments with RuWO /SiO -Ca were performed at different temperatures
x
2
(160°C, 180°C and 200°C) and time profiles were recorded (Figure 4). All three time profiles
have a similar shape. In two parallel pathways, hexanamide (1) is hydrogenated towards
hexylamine (3) (C-O cleavage) and hexanol (2) (C-N cleavage). From the initial slopes, it is
deduced that amine formation is promoted over alcohol formation when increasing the
temperature. After this initial phase, a consecutive product is detected, viz. dihexylamine (4).
Although it can be formed from solely the primary amine, dihexylamine (4) is mainly the
result of a condensation reaction between hexanol (2) and hexylamine (3) since this reaction
is much faster (Error! Reference source not found.). To be more precise, both the amine
and alcohol are at constant equilibrium with trace amounts of their corresponding imine and
aldehyde. These unsaturated molecules will react with amines, and after NH elimination and
3
reduction, secondary or even tertiary amines are formed (as for Ru/C). Hexanol (2) can be
transformed into hexylamine (3) as well (after reaction with NH ); at 160°C the rates of C-O
3
and C-N cleavage are roughly the same. Nevertheless, after a reaction time of 5h, the
hexanol (2) yield reaches a maximum while the amount of hexylamine (3) still increases.
These observations gave rise to the reaction network depicted in Figure 5. For all the
screened reaction temperatures, a maximum in hexylamine (3) yield was observed at 90-
92% conversion, with 50 bar H , 0.5 bar NH , in CPME. The hydrogenation experiment at
2 3
200°C resulted in the highest yield of 50% hexylamine (3) (90% conversion). We did not
perform hydrogenation experiments above 200°C, regarding the higher energy costs and
increased rate of undesired side reactions, e.g. amine-imine condensation and deamination
(at 200°C, <1% hexane yield at full conversion).

Page 7 of 16
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01310E
Figure 4 – Time profiles of hydrogenation of hexanamide at different temperatures: (top) 200°C, (middle) 180°C and (bottom)
2+
1
60°C. Reaction conditions: hexanamide (1 mmol), 50 bar H
2
; 0.5 bar NH
3
, 5 mol% Ru (RuWO
x
/SiO
2
-Ca ), undecane (20 µL),
CPME (10 mL).
O
OH
NH2
+
H2
-H O
2
NH2
NH
1
+
H2
-NH3
+
2
NH3
-
H O
O
NH2
+H2
3
+
H2
-NH3
+H2
-
H O
2
OH
N
2
H
4
x
Figure 5 – Reaction network of the hydrogenation of hexanamide with RuWO -catalyst.

Green Chemistry
Page 8 of 16
View Article Online
DOI: 10.1039/C9GC01310E
In order to further increase the primary amine yield, the ammonia and hydrogen pressure
were optimized. As one might expect, the addition of ammonia increases the selectivity for
hexylamine (3), since more alcohol is converted to hexylamine (3) and secondary amine
formation is hindered. On the other hand, compared to the amide, ammonia is preferentially
2+
adsorbed on the catalyst and therefore it lowers the catalytic activity of RuWO /SiO -Ca .
x
2
Thus, there is a clear trade-off between selectivity and activity (Figure 6). Assuming a first
order reaction (Error! Reference source not found.), we attempted to determine the
reaction order in [NH ] (Error! Reference source not found. – S 11). It appears that the
3
relation between ln(v ) and ln([NH ]) is not linear. At low concentrations of ammonia (P
NH3
init
3
between 0.5 – 1 bar) the reaction order is -0.3, whereas at higher concentrations of ammonia
PNH3 > 2 bar) the reaction order is approximately -0.9. We expect that the influence of amine
(
products is similar to the influence of ammonia. However, the amine concentration is very
low in comparison with the concentrations of ammonia. Therefore, the inhibiting effect of
amine products is probably limited.
Since it was our goal to obtain the highest possible primary amine yield under safe and eco-
friendly conditions, we selected 6 bar NH (= 4.5 M of ammonia in CPME at room
3
temperature, determined by weight) as ideal, with 77% hexylamine (3) yield (88%
conversion, 88% selectivity). This, to our knowledge, is already better than the best available
results for bimetallic catalysts described in literature, as described by Beamson et al. (2010)
and Hirosawa et al. (1996).12,20 In an industrial set-up, controlled addition of ammonia is a
powerful tool to steer towards the desired ratio of primary vs secondary amines, since
secondary amines are in some cases even more valuable than primary amines, e.g.
dimethylamine.39 Excess of ammonia can be easily recycled, e.g. by degassing and cooling
to liquid ammonia.
Whereas ammonia is applied to steer the selectivity, hydrogen mostly has an influence on
the hydrogenation rate (combined with 6 bar NH ). Error! Reference source not found. in
3
the Supporting Information shows that 25 bar H2 suffices to obtain an acceptable
hydrogenation rate.39

Page 9 of 16
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01310E
Figure 6 – Variation of the ammonia pressure in hydrogenation experiments of hexanamide: (left) fixed reaction time of 2h;
(
2
right) variable reaction time, with ~90% conversion. Reaction conditions: hexanamide (1 mmol), 200°C, 50 bar H , 5 mol% Ru
2+
x 2
(RuWO /SiO -Ca ), undecane (20 µL), CPME (10 mL).
With 6 bar of ammonia to control the primary amine selectivity, the contribution of a basic
catalyst support fades away. In Figure 7, RuWO on fumed SiO is compared with basic
x
2
cation loaded SiO and spinel (MgAl O , CaAl O ) supports. Whereas RuWO /SiO has a
2
2
4
2
4
x
2
hexylamine (3) selectivity of 84%, a selectivity of 85-90% is observed for RuWO on a
x
basified support (at ~90% conversion). Even more interesting is the behaviour of
RuWO /spinel. Wielding a spinel support drastically increased the catalytic activity, with 83%
x
-4
-1
hexylamine (3) yield at full conversion for RuWO /MgAl O (initial TOF of 2.2*10 s ). To
x
2
4
investigate this difference in reaction rate, RuWO /SiO and RuWO /MgAl O both were
x
2
x
2
4
characterized with N -physisorption and TEM(-EDX) (see Supporting Information). It
2
appeared that, while RuWO /SiO has the highest BET surface area (395 m²/g, vs 159 m²/g
x
2
in case of RuWO /MgAl O ), RuWO /MgAl O has the smallest Ru particle size, resulting in a
x
2
4
x
2
4
higher catalytic activity due to an increased specific metal surface (Ru particle diameter of
.5 - 2 nm vs 2 - 3 nm for RuWO /SiO ). A similar effect was observed when comparing the
1
x
2
performance of Ru/C with Ru/SiO (Figure 1, < 2nm vs 2 - 3 nm for Ru/SiO ). The TEM-EDX
2
2
data also show that W is very evenly distributed over the entire surface, not making clusters
like Ru does.

Green Chemistry
Page 10 of 16
View Article Online
DOI: 10.1039/C9GC01310E
Figure 7 – Catalytic screening of RuWO
x
on various supports: (left) SiO
2
vs Na-, Ca-, K- & Cs-loaded SiO
2
(7 h); (right) Mg- vs
Ca-spinel (4 h & 7 h). Reaction conditions: hexanamide (1 mmol), 200°C, 50 bar H
CPME (10 mL).
2 3
, 6 bar NH , 7 mol% Ru, undecane (20 µL),
To investigate the stability of RuWO /MgAl O , a recycling test was performed along with ICP
x
2
4
analysis and XRD of both fresh and used catalyst. The recycling test showed that
RuWO /MgAl O can be used at least 5 times without losing any catalytic activity nor
x
2
4
selectivity (the same goes for RuWO /SiO , Error! Reference source not found.); after
x
2
each run, a nearly full conversion with a selectivity for the primary amine of approximately
0% was observed. This is a rather interesting observation since literature states that the
8
selectivity of Beamson’s Ru-Mo catalyst for the primary amine drops significantly after three
runs.12 The high stability of this RuWO -catalyst could be confirmed by ICP; both Ru and W
x
content in the mixture after reaction were below the detection limits (1 ppb), which represents
a leaching below 0.005% and 0.02% respectively. The recorded XRD patterns do not show
any Ru peaks and essentially remain the same, once again illustrating the high stability of the
catalyst (no sintering, Ru nanoparticles are too small to be detected by XRD, Error!
Reference source not found.). Since the catalyst does not show signs of deactivation, we
are unable to calculate a turnover number (TON) until catalyst deactivation. However, a
complete conversion can be reached with even just 0.5 mol% of Ru, which means that at the
end of a single run, the TON amounts to at least 200.
A last key factor that requires investigation, is the (Lewis) acidity of the catalyst. It has been
proposed that these bimetallic compounds are excellent catalysts for the hydrogenation of
amides because the metal oxide promotor presents Lewis acid sites. These Lewis acids sites
help with the adsorption of amides to the surfaces. As a result, the interface PGM-metal
oxide acquires a high catalytic activity. To investigate the (Lewis) acidity of the catalyst, NH3-
TPD was performed. The data clearly show that there are two types of (Lewis) acid sites on
the catalyst’s surface (Figure 8). Based on the cumulative amount of NH adsorbed to the
3

Page 11 of 16
Green Chemistry
View Article Online
surface, we calculated that there are 0.055 mmol acid sites per g of catalyst. This is
DOI: 10.1039/C9GC01310E
approximately the same as the amount of tungsten per g of catalyst. It is logical to assign
these acid sites to the tungsten species, since the catalytic support is typically basic
(
MgAl O spinel). The data thus suggest a very high, nearly monoatomic dispersion of
2 4
tungsten. This was also suspected after analysing the TEM data. In literature, similar
observations have been made for Mo, which is chemically similar to W.44
Figure 8 – NH
3
-TPD with RuWO
x
/MgAl
2
O
4
-catalyst. After a phase of NH
3
adsorption 100°C and flushing the catalyst with pure
exhaust was monitored.
N
2
, the catalyst bed was gradually heated (5°C/min) and the actual desorption takes place. The NH
3
Finally, a substrate scope investigation was performed to illustrate the range of applicability
Table 1 & supporting information). High yields of the corresponding primary amines were
(
obtained for all aliphatic monofunctional primary amides (entries 1–5). The high yields and
selectivities are consistently obtained for chain lengths between C3 and C12. Hydrogenation
experiments with primary α,ω-diamides produced some interesting results; short bifunctional
molecules tend to follow a deamination/decarbonylation pathway resulting in monofunctional
primary amides, which react further as expected (entry 6). In case of longer α,ω-diamides,
each amide group reacts more independently. However, once an ‘amino-amide’ of
appropriate chain length is formed, the molecule easily performs an intramolecular
transamination resulting in a cyclic secondary amide. Only small amounts of the α,ω-diamine
can be detected. This explains why the results for adipamide and caprolactam are virtually
40
the same, with a yield of 85% azepane. In these two reactions, hexylamine is the main side
product, presumably formed after hydrogenolysis of the azepane ring. Such a cyclization
reaction was not observed for the C10 diamide, viz. decanediamide (entry 9). Acyclic
secondary and tertiary amides with very short N-substituents were reduced as well, although
this probably does not proceed via a direct hydrogenation of the secondary/tertiary amides
themselves (entries 10 & 12); rather ammonolysis will take place first, yielding primary
amides which react further to primary amines. The short amines (e.g. methyl-,
dimethylamine) were not detected due to their very low boiling points (e.g. -6°C and 7°C

Green Chemistry
Page 12 of 16
View Article Online
respectively). If the N-substituent is more bulky, the ammonolysis does not proceed, leading
DOI: 10.1039/C9GC01310E
to a low hydrogenation rate (entry 11). More activated tertiary amides such as N-
acetylmorpholine, can be easily hydrogenated as well (entry 13). Such molecules have an
increased preference for a C-N cleavage over C-O cleavage (45% morpholine yield). At
200°C, however, the morpholine ring itself isn’t stable. Besides morpholine, diethylamine and
some minor side product, the hydrogenation of N-acetylmorpholine resulted in a large
amount of ethanol and ethylamine (see supporting information).
x 2 4
Table 1 – Substrate scope investigation with RuWO /MgAl O .
O
5
mol% Ru (RuWO /MgAl O )
x 2 4
R'
R'
R
N
R
N
R''
1 mmol, 10 mL CPME, 6h
00°C, 6 bar NH + 50 bar H2
R''
2
3
X a
S b
[%]
S b
[%]
Substrate
Major product
Minor product
[%]
O
H
N
NH2
1
2
3
4
5
6
7
96
84
83
81
82
80
30
85
85
15
17
17
14
18
23c
8
NH2
O
> 99
99
NH2
NH2
HN
NH2
O
NH
NH2
NH
O
97
NH2
NH2
O
NH2
N
NH2
> 99
> 99
> 99
> 99
> 90d
93e
H
O
O
NH2
NH2
H2N
O
NH2
H
N
NH2
H2N
NH2
NH2
O
H
H
N
O
N
8
6
O
H N
H2N
2
main product d:
NH2
9
NH2
> 50%
O
O
H
N
NH2
1
0
1
2
3
87
13
/
N
H
O
1
1
NH
6
NH2
> 99
88
/
O
H
N
57e
9
NH2
N
O
HN
O
H
N
1
N
> 99
45
17
O
a
b
c
Conversion. Selectivity. Very volatile compound; therefore, yields are underestimated by
d
analysing the liquid phase. 1,10-decanediamine partially precipitates out of CPME resulting
e
in an less precise product yield calculation. Mixture of original amide and propionamide.

Page 13 of 16
Green Chemistry
View Article Online
Additionally, we discovered that secondary amines are not necessarily dead-end products.
DOI: 10.1039/C9GC01310E
With nearly any commercial PGM supported on carbon, we were able to cleave dihexylamine
4) into hexylamine (3) with 62% yield (Ru & Pt on carbon supports, Figure 10). The reverse
(
reaction with hexylamine (3) showed that this is the equilibrium position between mono- and
dihexylamine at 200°C, 10 bar H , 6 bar NH (= 4.5 M of ammonia) in CPME (Figure 9 &
2
3
Error! Reference source not found.). The capability to valorize secondary(/tertiary) amines
opens up many possibilities. First, 100% yield of the primary amine is in theory possible.
However, this would require a continuous isolation of the primary amine from the secondary
amine. Secondly, one can opt to perform the difficult amide hydrogenation in the absence of
ammonia to selectively obtain the secondary amine (Figure 6) and perform an ammonolysis
afterwards. Although a one-step hydrogenation process provides higher yields, the two-step
process would drastically increase the hydrogenation rate and replace a single
hydrogenation process at elevated pressure, by a combined process operating under milder
conditions. To our knowledge, the ammonolysis of secondary or tertiary amines has only
been reported once, viz. by Olin et al. in 1937.41 In their work, a yield of approximately 33%
monobutylamine could be obtained from dibutylamine with a MnO /C catalyst, operating at
x
292°C starting from a molar composition of 10-1 ammonia-dihexylamine.
N
H
-H2
N
+NH3
NH H2N
+H2
NH2
Figure 9 – Proposed reaction scheme of the
ammonolysis of secondary amines.
Figure 10 – Ammonolysis of dihexylamine with different commercial
PGM/C. Reaction conditions: hexanamide (1 mmol), 200°C, 10 bar H
bar NH , 5 mol% Ru, undecane (20 µL), CPME (10 mL), 5 h.
2
,
6
3
Conclusion
In conclusion, our new and relatively cheap ruthenium-tungsten bimetallic catalyst performed
excellently for the hydrogenation of aliphatic acyclic primary amides, such as
hexanamide (1). Reactions were conducted in CPME. Despite the fact that CPME slightly
lowers the selectivity for the primary amine, this benign and green alternative is a major
improvement when compared to the commonly used toxic solvents such as DME, dioxane
and hexane. It is also very stable, safe to work with and ensures an efficient product workup.

Green Chemistry
Page 14 of 16
View Article Online
To improve the primary amine selectivity, we explored the use of a strongly basic support
DOI: 10.1039/C9GC01310E
(
e.g. a spinel), and the use of NH partial pressure; the latter option is most effective. At
3
relatively high NH pressures, the additional effect of a basic support is noticeable but small.
3
After a further optimization, a yield of up to 83% of hexylamine (3) could be obtained with
RuWO /MgAl O at 200°C, with 50 bar H and 6 bar NH (= 4.5 M of ammonia), using
x
2
4
2
3
-4
-1
5
mol% Ru after 6 h in CPME (initial TOF of 2.2*10 s ). This green catalytic system is thus
a significant improvement in comparison with current technologies for the hydrogenation of
aliphatic acyclic primary amides. Finally, RuWO -catalysts were proven to be very robust and
x
applicable on a wide range of primary amides.
Acknowledgements
The authors appreciate the assistance of Karel Duerinckx, Free De Schouwer & Annelies
Vandekerkhove, Koen Adriaensen, Guangxia Fu, Cherry Cheung, Carlos Marquez, Thomas
1
Cuypers, Paul Van der Aerschot and Cédric Van Goethem with H NMR, pressure reactors,
spinel synthesis, N physisorption, NH temperature programmed desorption, ICP-OES
2
3
measurements, NH set-up, technical support and TEM-EDX imaging. We also thank the
3
financing of the Hercules fund (AKUL/13/19) and the generous support of Prof. Jin Won Seo
(Dep. MTM, KU Leuven) for TEM-EDX. R. C. is also very grateful for his SB PhD fellowship
at FWO.
Reference
(1) P. Roose, K. Eller, E. Henkes, R. Rossbacher, & H. Höke, Amines, Aliphatic. In Ullmann’s Encyclopedia of Industrial
Chemistry: Wiley-VCH Verlag, Germany, 2015; pp 1−55
(
(
(
(
(
(
(
(
(
(
(
2) S. A. Lawrence, Amines: Synthesis, Properties and Application: Cambridge University Press, 2004; pp 1-382
3) A. M. Smith & R. Whyman, Chemical Reviews, 2014, 114, 5477–5510.
4) B. Wojcik, H. Adkins, J. Am. Chem. Soc. 1934, 56, 2407−2419.
5) H. Adkins, Chem. Abstr. 1939, 33, 2906.
6) A. Guyer, A. Bieler, G. Gerliczy, Helv. Chim. Acta 1955, 38, 1649-1654.
7) T. Mitsudome, et al. Angewandte Chemie International Edition 2017, 56, 9381–9385.
8) T. Miura, I. E. Held, S. Oishi, M. Naruto & S. Saito, Tetrahedron Letters 2013, 54, 2674–2678.
9) Y. Nakagawa, R. Tamura, M. Tamura & K. Tomishige, Science and Technology of Advanced Materials 2015, 16, 14901.
10) M. Stein & B. Breit Angewandte Chemie International Edition 2013, 52, 2231–2234.
11) K. Shimizu, W. Onodera, A. S. Touchy, S. M. Siddiki, T. Toyao & K. Kon, ChemistrySelect 2016, 1, 736–740.
12) G. Beamson, A. J. Papworth, C. Philipps, A. M. Smith & R. Whyman, Advanced Synthesis and Catalysis 2010, 352, 869–
8
83.
13) Y. Takeda, T. Shoji, H. Watanabe, M. Tamura, Y. Nakagawa, K. Okumura & K. Tomishige, ChemSusChem, 2015, 8, 1170–
178.
(
1
(
(
(
(
(
(
14) M. Tamura, S. Ishikawa, M. Betchaku, Y. Nakagawa & K. Tomishige, Chemical Communications 2018, 54, 7503–7506.
15) M. W. Duch & A. M. Allgeier, Applied Catalysis A: General 2007, 318, 190–198.
16) G. Beamson, A. J. Papworth, C. Philipps, A. M. Smith & R. Whyman, Journal of Catalysis 2010, 269, 93–102.
17) G. Beamson, A. J. Papworth, C. Philipps, A. M. Smith & R. Whyman, Journal of Catalysis 2011, 278, 228–238.
18) R. Burch, et al. Journal of Catalysis 2011, 283, 89–97.
19) J. Coetzee, H. G. Manyar, C. Hardacre & D. J. Cole-Hamilton, ChemCatChem 2013, 5, 2843–2847

Page 15 of 16
Green Chemistry
(
(
(
(
(
(
(
(
(
(
20) C. Hirosawa, N. Wakasa & T. Fuchikami, Tetrahedron Letters 1996, 37, 6749–6752.
View Article Online
DOI: 10.1039/C9GC01310E
21) I. D. Dobson, European Patent EP 0286280 B1, 1988.
22) I. D. Dobson, US patent 4937384, 1990.
23) A. A. Smith, P. Dani, P. D. Higginson, A. J. Pettman, patent WO 2005/066112 A1, 2005.
24) C. J. Clarke, W.-C. Tu, O. Levers, A. Bröhl & J. P. Hallett, Chemical Reviews 2018, 118, 747–800.
25) W. Rupilius, Journal of Oil Palm Research 2011, 23, 1222–1226.
26) A. Corma Canos, S. Iborra & A. Velty, Chemical Reviews 2007, 107, 2411–2502.
27) J.-L. Dubois, A. Piccirilli, J. Magne, S. Tricoire, patent WO/2013/121151, 2013.
28) J.-P. Lallier, F. Mantisi, J.-P. Gillet, M. Hidalgo, patent WO/2013/088013 A1, 2013.
29) K. Visek, Amines, Fatty. In Kirk-Othmer Encyclopedia of Chemical Technology: John Wiley & Sons, Inc., Hoboken, NJ,
USA 2003, pp. 518–537.
(30) J. Becker, A. Lange, J. Fabarius & C. Wittmann, Current Opinion in Biotechnology 2015, 36, 168–175.
(31) M. Sjöblom, L. Matsakas, P. Christakopoulos & U. Rova, FEMS Microbiology Letters 2016, 363, 1–7.
(32) M. J. Campos Molina, R. Mariscal, M. Ojeda & M. López Granados, Bioresource Technology 2012, 126, 321–327.
(33) K. Watanabe, N. Yamagiwa & Y. Torisawa, Organic Process Research & Development 2007, 11, 251–258.
(34) F. M. Kerton, R. Marriott, J. Clark, G. A. Kraus, A. Stankiewicz, Y. Kou & P. Seidl, Alternative Solvents for Green Chemistry,
RSC Green Chemistry, 2013, pp 1-350.
35) D. Prat, A. Wells, J. Hayler, H. Sneddon, C. R. McElroy, S. Abou-Shehada & P. J. Dunn, Green Chemistry 2015, 18, 288–
96.
(
2
(
(
(
36) Sigma-Aldrich, Greener Solvent Alternatives (brochure), Sigma-Aldrich 2015, pp 1-4.
37) Evonik Resource Efficiency GmbH, Product information AEROSIL 380, Evonik Resource Efficiency GmbH 2018, pp 1-2.
38) J. P. Holmberg, E. Ahlberg, J. Bergenholtz, M. Hassellöv & Z. Abbas, Journal of Colloid and Interface Science 2013, 407,
168–176.
(
(
(
(
39) K. S. Hayes, Applied Catalysis A: General 2001, 221, 187–195.
40) R. Simonnot & S. Simmonot, patent FR1532777A, 1968.
41) J. F. Olin, T. E. Deger & I. M. Grosse, US patent 2192523, 1937.
42) T. Toyao, S. M. A. H. Siddiki, Y. Morita, T. Kamachi, A. S. Touchy, W. Onodera, K. Kon, S. Furukawa, H. Ariga, K. Asakura,
K. Yoshizawa & K.Shimizu, Chem. Eur. J. 2017, 23, 14848-14859.
43) J. Verduyckt, R. Coeck, D. E. De Vos, ACS Sustainable Chem. Eng. 2017, 54, 3290-3295.
44) T. Fransen, P.C. Van Berge & P. Mars, Studies in Surface Science and Catalysis, 1976,1, 405-420.
(
(

Green Chemistry
Page 16 of 16
View Article Online
DOI: 10.1039/C9GC01310E
Graphical abstract